Power Allocation System and Method Based on Distributed Consensus Algorithm

By generating an acyclic disturbance absorption map, the power deviation problem caused by changes in vehicle BMS parameters in the charging terminal scenario is solved, realizing the orderly acceptance and stable propagation of disturbance power, avoiding duplicate inclusion and grid-connected side impact, and improving the power distribution stability and efficiency of the charging station.

CN123121948APending Publication Date: 2026-11-17HENGCHU TECH (SHANXI) CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611270381.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-08-20
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

In multi-charging terminal scenarios, the vehicle BMS allows for power deviation caused by changes in charging parameters. Existing distributed consistent power allocation methods have problems such as uncontrolled disturbance propagation paths, easy duplication of disturbance power, easy concentration of receiving nodes, and abnormal adjustable power affecting the allocation results.

Method used

By generating an acyclic disturbance absorption map, the root node of the disturbance is determined and an absorption front is generated based on the disturbance hierarchy. This allows the charging terminal to take on the disturbance power layer by layer, and to reverse back the already taken disturbance power when the root node of the disturbance recovers. This controls the propagation path of the disturbance power and avoids repeated taking on and grid-connected side impact.

Benefits of technology

It reduces overcompensation and secondary load impact on the grid-connected side caused by parallel regulation of multiple terminals, and improves the stability and efficiency of power distribution.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123121948A_ABST
    Figure CN123121948A_ABST
Patent Text Reader

Abstract

This invention relates to the field of power allocation technology, and particularly to a power allocation system and method based on a distributed consensus algorithm. The invention proposes the following scheme: Obtaining the neighborhood communication relationships between multiple charging terminals, generating an original neighborhood communication graph, and obtaining the output power of each charging terminal and the vehicle's BMS-allowed charging parameters; when the output power of any charging terminal deviates from the currently allocated power due to changes in the vehicle's BMS-allowed charging parameters, determining the disturbance root node and generating an acyclic disturbance absorption graph; forming an absorption front based on the disturbance hierarchy, enabling charging terminals within the absorption front to absorb the disturbance power based on reliable and adjustable power, and transferring the unabsorbed disturbance power to the next absorption front; determining the target output power based on the absorption result, and reversing the absorbed disturbance power when the disturbance root node recovers, thereby reducing multi-path repeated absorption and grid-connected power surges.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of power allocation technology, and in particular to a power allocation system and method based on a distributed consensus algorithm. Background Technology

[0002] With the continuous increase in the number of new energy vehicles, public fast charging stations, multi-story parking garage fast charging stations, and centralized fast charging stations in industrial parks typically have multiple charging terminals to simultaneously meet the rapid charging needs of multiple vehicles. Although these multiple charging terminals connect to different vehicles and independently execute output control during operation, they are often collectively limited by the available power at the station level, the capacity of the distribution cabinet, the capacity of the distribution branch, or the capacity of the transformer connected to the system. Therefore, in multi-terminal parallel charging scenarios, how to allocate power among the charging terminals within the shared power supply capacity directly affects charging efficiency, equipment utilization, and grid-connected power fluctuations.

[0003] Existing power allocation methods for multiple charging terminals typically include centralized power scheduling and distributed consistent power allocation. Centralized power scheduling relies on the station-end control equipment to centrally collect the status of each charging terminal and uniformly issue power commands, which facilitates overall capacity constraint management. However, in scenarios with a large number of charging terminals, dispersed across floors or areas, and complex communication links, it places high demands on the real-time computing power, communication reliability, and completeness of global status collection of the station-end control equipment. Distributed consistent power allocation, on the other hand, allows each charging terminal to update its power based on its local status and the status of neighboring terminals, reducing dependence on a single centralized control node and making it suitable for fast charging scenarios with multi-area and multi-level deployments.

[0004] However, the vehicle charging process is not static. The vehicle's BMS dynamically adjusts the allowable charging current, allowable charging voltage, or requested charging power based on factors such as the battery's state of charge, battery temperature, individual cell voltage consistency, and protection strategies. When a charging terminal cannot continue to execute its current power allocation due to changes in the vehicle's BMS's allowable charging parameters, a finite power difference will form between its current allocated power and its executable output power. If this power difference is directly propagated in the original neighborhood communication relationship according to the normal neighborhood consistency update method, especially in mesh or ring neighborhood connections, the same power disturbance may be repeatedly sensed and compensated by different charging terminals through multiple paths, causing the received power to exceed the actual released power, thereby causing local branch load bias, power oscillation, and secondary impacts on the grid side.

[0005] Furthermore, when multiple charging terminals within the same neighborhood have a certain adjustable power, consistently using a fixed order, fixed ratio, or maximum adjustable power priority method to allocate disturbance power can easily lead to some charging terminals repeatedly prioritizing power during consecutive disturbance events, resulting in a concentrated local power adjustment burden. If the adjustable power of individual charging terminals deviates from its actual executable capability due to BMS communication lag, sampling errors, or local control anomalies, it may also lead to the incorrect allocation of disturbance power, requiring repeated compensation in subsequent levels. Therefore, existing distributed power allocation methods still suffer from problems such as uncontrolled disturbance propagation paths, the easy duplication of limited disturbance power, the tendency for receiving nodes to concentrate, and the impact of abnormal adjustable power on allocation results when dealing with local power deviations caused by sudden changes in vehicle-side charging constraints.

[0006] To address the above issues, this application presents a power allocation system and method based on a distributed consensus algorithm. Summary of the Invention

[0007] To address the shortcomings of existing technologies, this invention provides a power allocation system and method based on a distributed consensus algorithm. The system acquires the neighborhood communication relationships between multiple charging terminals, generates an original neighborhood communication graph, and obtains the output power of each charging terminal and the vehicle's BMS-permitted charging parameters. When the output power of any charging terminal deviates from the currently allocated power due to changes in the vehicle's BMS-permitted charging parameters, a disturbance root node is determined, and an acyclic disturbance absorption graph is generated. An absorption front is formed based on the disturbance hierarchy, enabling charging terminals within the absorption front to absorb the disturbance power based on reliable and adjustable power, and transferring any unabsorbed disturbance power to the next absorption front. The target output power is determined based on the absorption results, and the absorbed disturbance power is reversed when the disturbance root node recovers, thereby reducing multi-path repeated absorption and grid-connected power surges.

[0008] To achieve the above objectives, the present invention provides the following technical solution:

[0009] A power allocation method based on a distributed consensus algorithm, the method comprising:

[0010] Obtain the neighborhood communication relationship between multiple charging terminals within the charging station, generate the original neighborhood communication graph, and obtain the output power of each charging terminal and the vehicle BMS allowed charging parameters;

[0011] When the vehicle BMS of any charging terminal causes its output power to deviate from the currently allocated power due to changes in the charging parameters, the corresponding charging terminal is identified as the disturbance root node, and the downstream node is determined based on the disturbance root node in the original neighborhood communication graph to generate an acyclic disturbance absorption graph.

[0012] In the acyclic disturbance absorption diagram, an absorption front is generated according to the disturbance level, so that the charging terminal corresponding to the node located in the absorption front can accept the disturbance power based on its own adjustable power, and transfer the unaccepted disturbance power to the next level charging terminal.

[0013] The target output power is determined based on the absorption result of the disturbance power by each charging terminal, and when the output power of the disturbance root node is restored, the absorbed disturbance power is retracted in the opposite direction of the acyclic disturbance absorption diagram.

[0014] The neighborhood communication relationship is used to characterize whether there is a neighborhood state interaction relationship between any two charging terminals. The methods for obtaining the neighborhood communication relationship include:

[0015] By acquiring at least one of the power distribution access identifier, installation location identifier, and communication connection status of each charging terminal, two charging terminals that meet the preset adjacent conditions are identified as having a neighborhood status interaction relationship. The preset adjacent conditions include at least one of the following: connected to the same power distribution branch, connected to the same power distribution cabinet, located in the same charging area, located in adjacent charging areas, or having a direct communication link.

[0016] The original neighborhood communication graph is constructed in the following ways:

[0017] Each charging terminal is used as a graph node. Based on the neighborhood communication relationship, a connection edge is established between two graph nodes with a neighborhood state interaction relationship. An adjacency matrix is ​​generated based on each graph node and its corresponding connection edge.

[0018] The original neighborhood communication graph is generated based on the adjacency matrix, wherein the original neighborhood communication graph is used to characterize the node connection relationship that allows each charging terminal to directly interact with the state during the power allocation process.

[0019] The vehicle BMS allowed charging parameters are used to characterize the charging constraints that the vehicle battery management system allows the charging terminal to output in the current charging state. The vehicle BMS allowed charging parameters include allowed charging current, allowed charging voltage and requested charging power.

[0020] The vehicle BMS allows charging parameters to be obtained through the charging communication link between the charging terminal and the vehicle battery management system. The charging communication link includes at least one of the following: the vehicle communication interface after the charging gun is connected, the charging controller communication interface, or the charging protocol message exchange link.

[0021] The generation method of the acyclic perturbation absorption map includes:

[0022] The disturbance root node is taken as the currently included node, and the charging terminal that has a neighborhood state interaction relationship with the currently included node and has not yet been included in the current power disturbance is determined as the candidate node.

[0023] Determine whether the candidate node meets the disturbance absorption entry condition. The disturbance absorption entry condition includes: the charging terminal corresponding to the candidate node has an adjustable power that matches the power disturbance direction, and the adjustable power is greater than or equal to a preset power threshold.

[0024] Candidate nodes that meet the disturbance absorption entry conditions are included in the current power disturbance, and the candidate nodes included in the current power disturbance are determined as the downstream nodes of the currently included nodes;

[0025] Perform iterative operations, taking downstream nodes as currently included nodes, and repeat the inclusion operation for downstream nodes until there are no more candidate nodes to be included.

[0026] The acyclic perturbation absorption map is generated based on the included nodes and their connection relationships.

[0027] The absorption front is generated in the following ways:

[0028] Based on the iteration round corresponding to when each included node is included in the current power disturbance, the disturbance level of each included node relative to the disturbance root node is determined, wherein nodes included in the same iteration round belong to the same disturbance level;

[0029] The group of included nodes that has not yet completed perturbation power processing and has the smallest perturbation level is identified as the current absorption frontier.

[0030] The method further includes:

[0031] The charging terminal corresponding to the node within the current absorption front edge is controlled to accept the received disturbance power, and when there is unaccepted disturbance power in the charging terminal within the current absorption front edge, the node included in the next disturbance level is updated as a new absorption front edge.

[0032] The update of the absorption front is stopped when the disturbance power is fully absorbed or there are no nodes included in the next disturbance level.

[0033] The methods for receiving the disturbance power include:

[0034] Determine the disturbance power to be received corresponding to the current absorption front. The disturbance power to be received corresponding to the first absorption front is determined based on the change in output power of the disturbance root node, and the disturbance power to be received corresponding to subsequent absorption fronts is determined based on the disturbance power that was not received by the previous absorption front.

[0035] The adjustable power of each charging terminal within the current absorption front is determined respectively, and the disturbance power acceptance share of each charging terminal is determined according to the adjustable power of each charging terminal and the disturbance power to be accepted. The sum of the disturbance power acceptance shares is less than or equal to the disturbance power to be accepted, and the disturbance power acceptance share of any charging terminal is less than or equal to its corresponding adjustable power.

[0036] The output power of each charging terminal is adjusted according to the corresponding disturbance power acceptance share, and the unaccepted disturbance power to be transmitted to the next absorption front is determined based on the difference between the disturbance power to be accepted and the actual power accepted by each charging terminal.

[0037] The method for determining the share of disturbance power absorption includes:

[0038] Based on the adjustable power of each charging terminal within the current absorption frontier, the vehicle BMS allowed charging parameters, and historical acceptance results, the adjustable power of each charging terminal is reliably verified to obtain the reliable adjustable power corresponding to each charging terminal.

[0039] Charging terminals with reliable adjustable power greater than or equal to the preset acceptance lower limit are identified as candidate acceptance nodes.

[0040] Based on the disturbance event identifier, disturbance level, disturbance root node identifier, and node identifier of each candidate node corresponding to the current power disturbance, a succession ranking factor is generated for each candidate node, and the succession order of each candidate node is determined according to the succession ranking factor.

[0041] According to the acceptance order, the disturbance power acceptance share of each candidate node is determined sequentially within the range of the disturbance power to be accepted, wherein the disturbance power acceptance share of any candidate node is less than or equal to its corresponding reliable adjustable power, and the sum of the disturbance power acceptance shares is less than or equal to the disturbance power to be accepted.

[0042] A power allocation system based on a distributed consensus algorithm, the system comprising:

[0043] The data acquisition module is used to acquire the neighborhood communication relationship between multiple charging terminals in the charging station, generate the original neighborhood communication map, and acquire the output power of each charging terminal and the vehicle BMS allowed charging parameters.

[0044] The disturbance absorption graph generation module is used to determine the corresponding charging terminal as the disturbance root node when the vehicle BMS of any charging terminal allows changes in charging parameters, and to determine the downstream node based on the disturbance root node in the original neighborhood communication graph to generate an acyclic disturbance absorption graph.

[0045] The disturbance power receiving module is used to generate an absorption front according to the disturbance level in the acyclic disturbance absorption diagram, so that the charging terminal corresponding to the node located in the absorption front receives the disturbance power based on its own adjustable power, and transfers the unreceived disturbance power to the next level charging terminal.

[0046] The target power determination module is used to determine the corresponding target output power based on the absorption result of the disturbance power by each charging terminal, and to back off the absorbed disturbance power in the opposite direction of the acyclic disturbance absorption diagram when the output power of the disturbance root node recovers.

[0047] Compared with the prior art, the beneficial effects of the present invention are:

[0048] When power deviations are caused by changes in the vehicle's BMS-permissible charging parameters, this application limits the propagation path of the disturbance power through an acyclic disturbance absorption map and progressively accepts the finite disturbance power based on the absorption front, avoiding the same power difference being repeatedly counted in the multi-path neighborhood relationship. Simultaneously, reliable verification of adjustable power and perturbation of the acceptance order can reduce local load concentration caused by abnormal declarations and fixed priorities, improving the stability of multi-terminal power allocation. Attached Figure Description

[0049] Other features, objects, and advantages of the invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0050] Figure 1 An exemplary application scenario diagram provided for an embodiment of this application;

[0051] Figure 2 A flowchart illustrating the power allocation method based on a distributed consensus algorithm provided in this application embodiment;

[0052] Figure 3 This is a schematic diagram of the structure of the original neighborhood communication graph provided in the embodiments of this application. Detailed Implementation

[0053] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0054] The term "embodiment" as used herein means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0055] This application's embodiments are applicable to charging station scenarios where multiple charging terminals operate in parallel and share station-level power supply capacity. They are particularly suitable for multi-level fast charging stations, centralized fast charging stations in parking garages, multi-level urban public fast charging stations, park fast charging stations equipped with multiple DC charging terminals, and charging systems employing split power cabinets and collaborative output from multiple charging terminals. In these scenarios, each charging terminal typically connects to different vehicles and independently executes output control. The terminal itself can adjust its output voltage, current, and power based on the allowable charging parameters fed back by the vehicle's battery management system. However, from the perspective of station-level power supply, the multiple charging terminals do not operate completely in isolation but are collectively constrained by the capacity of the distribution cabinet, the capacity of the distribution branch, the transformer access capacity, or the station-level available power limit. In other words, each charging terminal is independent at the control execution level but shares power resources; a decrease or recovery in the power of one terminal will affect the power adjustment space available to other terminals.

[0056] In actual fast charging, the charging demand on the vehicle side is not constant. As the vehicle's state of charge, battery temperature, individual cell voltage consistency, and the current-limiting strategy of the vehicle's battery management system change, the vehicle's BMS-allowed charging current, allowed charging voltage, or requested charging power will change during charging. When a vehicle enters a power reduction phase, triggers temperature protection, or reduces its requested power due to BMS control strategies, the corresponding charging terminal, even if it maintains its original allocation target, may be unable to continue outputting its original allocated power. At this time, the power originally allocated to that charging terminal but which it can no longer execute is not an arbitrarily amplified scheduling quantity, but rather a finite power difference formed between the charging terminal's current allocated power and the executable output power constrained by the vehicle's BMS. This finite power difference can be absorbed by other charging terminals that still have the capacity to accept the load, thereby improving the overall power utilization of the charging station. However, this absorption process must be subject to the combined constraints of station-level shared capacity and vehicle-side capacity constraints.

[0057] Existing charging station power allocation typically employs either centralized control or distributed consensus control. Centralized control involves a station-level controller uniformly collecting the status of each charging terminal and vehicle, then calculating the target output power for each terminal. Its advantage is a relatively complete global perspective; however, in multi-level fast charging stations or large-scale multi-terminal scenarios, it is easily limited by communication links, controller computational load, and the risk of single-point failures. Distributed consensus control, on the other hand, uses state interactions between adjacent charging terminals to gradually form power allocation results without a centralized controller or under weakly centralized conditions. This is more suitable for charging stations with a large number of terminals, complex spatial distribution, and localized communication links. However, when the vehicle's BMS allows for sudden changes in charging parameters, ordinary distributed consensus control cannot simply adopt the neighboring power difference update logic under normal operating conditions.

[0058] Specifically, the limited power difference released by a charging terminal due to vehicle-side limitations may be simultaneously sensed by multiple neighboring nodes through different communication paths. If each node still uses the ordinary consensus algorithm to treat this power change as a general power deviation in the update, the same limited power difference may be repeatedly counted through multiple paths. This causes multiple charging terminals to believe that there is available power space and simultaneously increase their output power in the same or adjacent control cycles. This may not only cause the actual power increase of the entire station to exceed the originally released power difference, but may also trigger a network-wide redistribution when the vehicle charging capacity of the disturbed root node recovers, resulting in overcompensation, reverse callback, and low-frequency oscillations, which in turn aggravates the load fluctuations on the grid-connected side.

[0059] Based on the above understanding, the embodiments of this application do not simply regard changes in vehicle BMS parameters as ordinary power errors, nor do they directly equate the communication relationship between adjacent terminals with the propagation path of disturbance power. Instead, they treat the power difference caused by changes in vehicle-side constraints as a power disturbance with a source and direction. For normal, small, and continuous power differences, each charging terminal can still perform distributed consistent power adjustment based on the original neighborhood communication graph. For deviations in executable output power caused by changes in the allowable charging parameters of the vehicle BMS, the charging terminal with the deviation is taken as the root node of the disturbance, and a loop-free disturbance absorption graph is temporarily formed in the original neighborhood communication graph. This prevents the finite power difference from freely spreading along the original mesh communication relationship, but instead allows it to be received or transferred step by step along a controlled hierarchical direction.

[0060] It should be noted that the acyclic disturbance absorption graph described in this application is not intended to re-establish ordinary communication links, nor does it require a master-slave control relationship between charging terminals. Each charging terminal still independently executes its own output control, and determines its target output power solely based on its own vehicle BMS allowed charging parameters, current output power, and neighborhood interaction results. The purpose of the acyclic disturbance absorption graph is to temporarily constrain the propagation range and direction of the same finite disturbance power under shared power capacity constraints, preventing the disturbance power from being repeatedly propagated, repeatedly received, or amplified back in the multi-path structure of the original neighborhood communication graph. When the output power of the disturbance root node recovers, the disturbance power that has been received by other charging terminals can gradually retreat along the opposite direction of the acyclic disturbance absorption graph, thereby avoiding triggering re-competition and synchronization adjustment across the entire station during the recovery phase.

[0061] Therefore, the technical problem solved by the embodiments of this application can be understood as follows: In a distributed fast charging scenario where multiple charging terminals independently perform output control but share station-level power capacity, how to ensure that the limited disturbance power generated by the vehicle BMS is received in a one-time and orderly manner in the neighborhood communication network when the executable output power of a certain terminal deviates from the current allocated power due to changes in the charging parameters allowed by the vehicle BMS, and suppress the disturbance power from being repeatedly included in the consistency update in the multi-path topology, thereby reducing the overcompensation and secondary load impact on the grid-connected side caused by the parallel adjustment of multiple terminals.

[0062] refer to Figure 1 , Figure 1 This is an exemplary application scenario diagram provided for an embodiment of this application.

[0063] like Figure 1 As shown, the application scenario can include a processor, controller, power distribution terminal, and multiple charging terminals. These charging terminals can be designated as Charging Terminal 1, Charging Terminal 2, Charging Terminal 3, and so on, up to Charging Terminal N, where N is an integer greater than or equal to 3. The power distribution terminal can connect to a station-level power supply line, distribution cabinet, power cabinet, or the low-voltage output of a transformer, providing power access to multiple charging terminals. This allows multiple charging terminals to share the station-level available power, distribution branch capacity, or power cabinet output capacity at the electrical level. Each charging terminal can connect to the vehicle to be charged and obtain the corresponding vehicle's BMS-permitted charging parameters through the charging gun communication interface, charging controller communication interface, or charging protocol message exchange link, enabling it to independently execute its output control.

[0064] The processor can be used to run the power distribution control program provided in the embodiments of this application, and the controller can be used to coordinate data interaction and power output control between the power distribution terminal and multiple charging terminals based on the calculation results of the processor or the local calculation results of each charging terminal. In some embodiments, the processor and controller can be located in the station-level control equipment of the charging station; in other embodiments, the functions of the processor and controller can be distributed in the power distribution terminal or the local control unit of each charging terminal. In other words, Figure 1 The processor and controller shown are not limited to being the only control center that must perform centralized power distribution to all charging terminals; they can serve as a computing execution unit, communication management unit, or control command carrying unit in the distributed power distribution process.

[0065] During operation, each charging terminal can periodically collect its own output power, current allocated power, and vehicle BMS-permitted charging parameters, and form neighborhood state interactions based on the power distribution access relationship, installation location relationship, or communication connection relationship between charging terminals. This neighborhood state interaction can be achieved through direct communication links between charging terminals, or through message forwarding by the controller or power distribution terminal. Regardless of the communication method used, each charging terminal participating in power allocation can update its own target output power based only on its own terminal state and the states of neighboring terminals, without requiring any charging terminal to continuously obtain the complete operating status of all charging terminals in the entire station.

[0066] When multiple charging terminals are charging in parallel, although each terminal connects to different vehicles and independently controls its output, they share the limited power supply capacity provided by the power distribution terminal. For example, a charging terminal may have been allocated a higher output power in the previous power allocation cycle, but if the connected vehicle experiences a decrease in its BMS-allowed charging parameters due to increased state of charge, battery temperature protection, or BMS current limiting strategy, the charging terminal may be unable to continue executing its current power allocation. In this case, a limited disturbance power will form between the current allocated power of the charging terminal and the executable output power constrained by the vehicle's BMS. This disturbance power corresponds to the power difference within the station that can be absorbed by other charging terminals with the capacity to accept the load, rather than scheduling resources that can be arbitrarily reused or amplified.

[0067] Next, with reference to the accompanying drawings, the power allocation method based on a distributed consensus algorithm provided in the embodiments of this application will be further described. Figure 2 The methods shown include:

[0068] S1: Obtain the neighborhood communication relationship between multiple charging terminals in the charging station, generate the original neighborhood communication graph, and obtain the output power of each charging terminal and the vehicle BMS allowed charging parameters;

[0069] In this embodiment, the charging station can be a multi-level fast charging station, a parking garage centralized fast charging station, or a DC fast charging station with multiple power modules connected in parallel. Multiple charging terminals are connected to different vehicles and independently perform output control, but the multiple charging terminals are all limited by the station-level available power, power distribution branch capacity, or power cabinet output capacity provided by the power distribution terminal.

[0070] S2: When the vehicle BMS of any charging terminal changes its allowed charging parameters, causing its output power to deviate from the currently allocated power, the corresponding charging terminal is identified as the disturbance root node, and the downstream node is determined based on the disturbance root node in the original neighborhood communication graph to generate an acyclic disturbance absorption graph.

[0071] In this embodiment, the vehicle BMS does not trigger disturbance processing simply because there is a fluctuation in the charging parameters. Instead, it uses a substantial deviation of the executable output power of the corresponding charging terminal from the currently allocated power as the criterion for judgment.

[0072] For example, a charging terminal's current allocated power in the previous allocation cycle was 80kW, but due to an increase in the vehicle's state of charge or battery temperature protection causing a decrease in the BMS's allowed charging current, the corresponding charging terminal can only stably output 50kW in the current cycle. This creates a finite disturbance power of 30kW. This disturbance power is not a power resource that can be arbitrarily amplified, but rather the power difference originally allocated to the disturbance root node that cannot be executed under vehicle-side constraints.

[0073] S3: In the acyclic disturbance absorption diagram, an absorption front is generated according to the disturbance level, so that the charging terminal corresponding to the node located in the absorption front can accept the disturbance power based on its own adjustable power, and transfer the unaccepted disturbance power to the next level charging terminal.

[0074] In this embodiment, after the acyclic disturbance absorption diagram is formed, each included node can form a disturbance hierarchy according to its inclusion round relative to the disturbance root node. Lower disturbance levels prioritize handling disturbance power, while higher disturbance levels only participate in handling when the previous level node fails to fully absorb the disturbance power. The absorption front can be understood as a group of nodes at the same level currently handling the disturbance power. For example, the first-level node directly connected to the disturbance root node first determines its own vehicle BMS allowed charging parameters, current output power, rated output capacity, and adjustment step size for this cycle to determine the adjustable power that can increase the output. If the total adjustable power of the first-level nodes is sufficient to absorb all the disturbance power, the disturbance absorption ends at the first level, and subsequent levels do not need to take action. If there is still an unabsorbed portion, the remaining disturbance power is then passed to the next level along the acyclic disturbance absorption diagram. This setting is not simply about pursuing the fastest allocation, but rather about prioritizing the absorption of power differences within a local area close to the disturbance root node, avoiding indiscriminate participation of distant nodes in adjustment, and reducing distribution branch fluctuations caused by large-scale synchronous power increases.

[0075] S4: Determine the corresponding target output power based on the result of each charging terminal absorbing the disturbance power, and when the output power of the disturbance root node recovers, back off the absorbed disturbance power in the opposite direction of the acyclic disturbance absorption diagram.

[0076] In this embodiment, after each charging terminal completes the absorption of disturbance power, it can add the corresponding absorption share to the original target output power or deduct it from the original target output power to form the target output power for the current control cycle, and the corresponding charging terminal independently performs output adjustment. If a charging terminal obtains an absorption share but its actual output fails to follow, for example, if the vehicle's BMS subsequently limits current or the terminal's local adjustment is limited, the unabsorbed portion can continue to be passed on as remaining disturbance power to the next absorption front, avoiding discrepancies between the book absorption and the actual output. On the other hand, when the disturbance root node is connected to the vehicle's BMS and allows the charging parameters to recover, enabling the disturbance root node to recover its output power, it no longer initiates a full-range power competition through the original neighborhood communication graph. Instead, based on the absorption path in the previous acyclic disturbance absorption graph, it gradually backs away the disturbance power that has been absorbed by other nodes in the opposite direction. In other words, the charging terminal that previously absorbed the disturbance power relinquishes its corresponding share within its power reduction range, and the disturbance root node gradually recovers its output accordingly. By adopting this corresponding acceptance and retreat processing method, the problem of disturbance once when the vehicle-side power decreases and then disturbance again when the vehicle-side power recovers can be avoided. This makes the power difference transfer under shared capacity have a traceable direction and boundary, reducing overcompensation, reverse callback and secondary load impact on the grid-connected side when multiple charging terminals are running in parallel.

[0077] It should be noted that the specific method for determining the target output power and the method for backing up the already received disturbance power can be implemented using relatively mature charging power closed-loop control, power ramping control, power limiting control, and distributed consistency update methods in this field. For example, when determining the target output power, the target power of the charging terminal in the previous control cycle can be superimposed with the disturbance power received share in the current cycle, and the limiting processing can be performed in combination with the vehicle BMS allowed charging parameters, the rated output capacity of the charging terminal, and the maximum adjustment step size in a single cycle; when backing up the received disturbance power, the corresponding charging terminal can be controlled to gradually reduce the increased output power according to the received power record, the backoff step size, and the recovery requirements on the vehicle side, so that the disturbance root node can gradually restore output. The above-mentioned power command generation, limiting, ramping, and closed-loop following control are all conventional implementation methods in the field of charging control that can be configured according to the field control cycle, power module response speed, and charging protocol requirements, and this application does not impose further limitations on them.

[0078] Next, we will further elaborate on the technical content of the method of this application regarding the original neighborhood communication graph.

[0079] In one example, the neighborhood communication relationship is used to characterize whether there is a neighborhood state interaction relationship between any two charging terminals, and the method for obtaining the neighborhood communication relationship includes:

[0080] By acquiring at least one of the power distribution access identifier, installation location identifier, and communication connection status of each charging terminal, two charging terminals that meet the preset adjacent conditions are identified as having a neighborhood status interaction relationship. The preset adjacent conditions include at least one of the following: connected to the same power distribution branch, connected to the same power distribution cabinet, located in the same charging area, located in adjacent charging areas, or having a direct communication link.

[0081] Understandably, obtaining neighborhood communication relationships can be achieved using mature configuration reading or link identification methods developed during the construction, commissioning, and operation of charging stations. The purpose is to determine which charging terminals can interact with each other in a neighborhood state, rather than redesigning the physical structure or communication protocol of the charging terminals. For example, the power distribution access identifier corresponding to each charging terminal can be read from the station's configuration file, power distribution cabinet wiring table, power module allocation table, or charging terminal installation ledger to identify charging terminals connected to the same power distribution cabinet, the same power distribution unit, or the same power distribution branch. Alternatively, adjacent installation locations can be determined based on the charging terminal installation number, parking space number, floor number, area number, or construction layout diagram, thus identifying charging terminals located in the same or adjacent areas as terminals with neighborhood state interaction relationships. Furthermore, communication handshakes, heartbeat messages, link quality checks, or station-side communication routing tables between charging terminals can determine whether two charging terminals have the capability for direct or relayed state interaction via station-side equipment.

[0082] Those skilled in the art will understand that the aforementioned power distribution access identifier, installation location identifier, and communication connection status are all basic information that can typically be obtained during the construction and operation of charging stations, and their specific sources can be selected based on the type of on-site equipment and communication architecture. For example, for DC fast charging stations using split power cabinets, the neighborhood communication relationship can be determined primarily based on the power cabinet output circuit, charging terminal number, and power module association; for fast charging stations in multi-story parking garages, the neighborhood communication relationship can be determined by combining floor numbers, parking space group numbers, and area communication gateways; for charging stations with direct inter-terminal communication capabilities, the neighborhood communication relationship can be determined based on the communication handshake results and link status between terminals. It is sufficient to identify, at a minimum, the objects from which each charging terminal can directly exchange status information during distributed power allocation; this application does not impose further limitations in this regard.

[0083] In yet another example, the original neighborhood communication graph is constructed in the following ways:

[0084] Each charging terminal is used as a graph node. Based on the neighborhood communication relationship, a connection edge is established between two graph nodes with a neighborhood state interaction relationship. An adjacency matrix is ​​generated based on each graph node and its corresponding connection edge.

[0085] The original neighborhood communication graph is generated based on the adjacency matrix, wherein the original neighborhood communication graph is used to characterize the node connection relationship that allows each charging terminal to directly interact with the state during the power allocation process.

[0086] It is understandable that constructing the original neighborhood communication graph using graph nodes, connecting edges, and adjacency matrices is a relatively mature representation method in graph structure data modeling. Its role is to convert whether charging terminals can directly interact with each other into a data structure that can be read and traversed by the processor. For this application, the adjacency matrix is ​​not used to additionally limit the master-slave relationship between charging terminals, nor is it required to set complex weights or communication attributes in the connecting edges. It is only used to indicate whether there is a neighborhood state interaction relationship between any two charging terminals, thereby providing the basic connection relationship for the subsequent generation of distributed consistent power allocation and acyclic disturbance absorption graph.

[0087] refer to Figure 3 , Figure 3 This is a schematic diagram of the structure of the original neighborhood communication graph provided in the embodiments of this application.

[0088] like Figure 3 As shown, in practical implementation, each charging terminal in the charging station can first be numbered. For example, charging terminal one, charging terminal two, charging terminal three, and charging terminal four can be denoted as node V1, node V2, node V3, and node V4, respectively. Then, based on the aforementioned neighborhood communication relationship, it is determined whether any two nodes have a neighborhood state interaction relationship, and the determination result is written into the adjacency matrix. For example, the adjacency matrix can be represented as a two-dimensional array A, where the data in the i-th row and j-th column of A represents the connection state between node Vi and node Vj. When node Vi and node Vj have a neighborhood state interaction relationship, the corresponding data can be recorded as 1; when they do not have a neighborhood state interaction relationship, the corresponding data can be recorded as 0. For bidirectional state interaction relationships, the adjacency matrix can be a symmetric matrix; for the connection relationship between a node itself, the diagonal data can be set to 0 to avoid self-looping during subsequent disturbance propagation.

[0089] like Figure 3 As shown, if node V1 has neighboring state interaction relationships with nodes V2 and V3 respectively, node V2 has a neighboring state interaction relationship with node V4, and node V3 does not have a direct neighboring state interaction relationship with node V4, then the corresponding adjacency matrix A can be represented as:

[0090] ;

[0091] In the example above, a value of 1 in the first row and second column indicates that charging terminal one and charging terminal two can directly exchange neighborhood state information; a value of 1 in the first row and third column indicates that charging terminal one and charging terminal three can directly exchange neighborhood state information; and a value of 0 in the third row and fourth column indicates that charging terminal three and charging terminal four do not directly exchange neighborhood state information. Therefore, when performing a neighborhood query, the processor can quickly determine the neighboring nodes of a node by reading the matrix row or column corresponding to that node. For example, reading the matrix row corresponding to node V1 can determine that the neighboring nodes of node V1 include nodes V2 and V3.

[0092] It should be noted that the data format of the adjacency matrix can be adaptively adjusted according to the number of charging terminals and computing resources. For example, when the number of charging terminals is small, a regular two-dimensional array or Boolean matrix can be used to store the node connection relationships; when the number of charging terminals is large and the actual connection relationships are sparse, a sparse matrix, adjacency list, or key-value mapping table can also be used to store the same connection relationships. The above data structures are all mature implementation methods in graph structure construction. They only need to be able to express the direct state interaction relationships between each charging terminal to a minimum and support subsequent queries of the adjacent nodes of a certain node. This application does not impose further limitations on this.

[0093] Next, we will further elaborate on the technical content of the method of this application regarding the absorption map without a loop.

[0094] In one example, the vehicle BMS allowed charging parameters are used to characterize the charging constraints that the vehicle battery management system allows the charging terminal to output in the current charging state. The vehicle BMS allowed charging parameters include allowed charging current, allowed charging voltage, and requested charging power.

[0095] The vehicle BMS allows charging parameters to be obtained through the charging communication link between the charging terminal and the vehicle battery management system. The charging communication link includes at least one of the following: the vehicle communication interface after the charging gun is connected, the charging controller communication interface, or the charging protocol message exchange link.

[0096] Understandably, the acquisition of charging parameters by the vehicle's BMS can be achieved using a relatively mature charging communication interaction method between the charging terminal and the vehicle. The purpose is to allow the charging terminal to obtain the charging constraints acceptable to the vehicle at the current charging stage, rather than modifying the vehicle's BMS or the charging communication protocol itself. Typically, after the charging gun completes connection confirmation, the charging terminal and the vehicle's BMS enter a charging handshake, parameter negotiation, and charging status interaction process. The vehicle's BMS can feed back parameters such as allowed charging current, allowed charging voltage, and requested charging power to the charging terminal, which then determines the output range it can execute at the current moment.

[0097] For example, vehicle BMS-permitted charging parameters can be obtained through charging demand messages, charging status messages, or charging parameter negotiation messages in the charging protocol messages. The charging terminal's charging controller can parse the permitted charging current, permitted charging voltage, and requested charging power fields carried in the messages, and use the parsing results as vehicle-side charging constraints input into the subsequent power allocation process. Vehicle BMS-permitted charging parameters can also be obtained through the charging controller's internal protocol stack cache, communication interface registers, or runtime data cache. For instance, after receiving BMS messages from the vehicle communication interface in each control cycle, the charging controller writes the latest permitted charging current and permitted charging voltage into its local cache for use by the power allocation program. In charging scenarios with station-side communication management equipment, the station-side communication management equipment can also receive the BMS parameter parsing results uploaded by each charging terminal and then forward them to the charging terminals participating in neighborhood status interaction.

[0098] Those skilled in the art will understand that the specific acquisition link for the vehicle's BMS-permitted charging parameters can be configured according to the charging terminal type, vehicle communication protocol, charging controller architecture, and on-site communication method. For example, a DC fast charging terminal can obtain BMS messages through the vehicle communication interface connected to the charging gun; in a split power cabinet scenario, the terminal controller can collect vehicle-side messages and then send the parsed parameters to the power cabinet control unit; a charging terminal with local edge computing capabilities can also directly complete message parsing and parameter updates within its own terminal control board. As long as the charging constraints allowed by the vehicle side can be obtained to a minimum, enabling the charging terminal to determine whether the current power allocation is still feasible, this application does not impose further limitations.

[0099] It is important to emphasize that the method of this application is applied when changes in the vehicle BMS's allowed charging parameters of any charging terminal cause its output power to deviate from the currently allocated power. This deviation can be understood as follows: within the current power allocation cycle, there is a difference between the current allocated power that the charging terminal should execute based on the distributed consistency power allocation result and the executable output power that can be stably executed after being constrained by the vehicle BMS's allowed charging current, allowed charging voltage, or requested charging power, and this difference reaches a level that affects the redistribution of shared power capacity. Specifically, reaching a level that affects the redistribution of shared power capacity means that the difference has exceeded the allowable fluctuation range caused by power sampling errors, communication jitter, control cycle asynchrony, or transient responses of the power module, sufficient to cause adjustments to the target output power of other charging terminals. Specifically, the absolute value of the difference can be compared with a preset deviation threshold, and the duration or number of consecutive occurrences of the difference can be used to confirm whether it has reached a level that affects the redistribution of shared power capacity. The preset deviation threshold can be set based on the charging terminal's rated output power, minimum power adjustment resolution, maximum adjustment step size per cycle, normal operating power fluctuation range, and the charging station's allowable power allocation error. The above confirmation process can be implemented using threshold comparison, hysteresis determination, continuous period determination or sliding time window statistics methods well known to those skilled in the art, which will not be elaborated upon in this application.

[0100] For example, if a charging terminal is currently allocated 80kW, but the vehicle's BMS allows a reduction in charging current due to increased state of charge or battery temperature protection, causing the charging terminal to only stably output 50kW in the current control cycle, then the charging terminal's output power can be considered to have deviated from its current allocated power. The resulting 30kW deviation can be considered as the disturbance power to be absorbed. Similarly, if a charging terminal previously only output 50kW due to current limiting by the vehicle's BMS, but subsequently the vehicle's BMS allows the charging parameters to recover, enabling it to resume output at its original or new allocated power, then the charging terminal can be considered to have a recovery deviation. The disturbance power already absorbed by other charging terminals needs to be reversed along the opposite direction of the acyclic disturbance absorption diagram.

[0101] Those skilled in the art will understand that the deviation should not be interpreted as any instantaneous fluctuation in the vehicle's BMS-permissible charging parameters, nor as a minor difference caused by sampling errors, communication jitter, or short-term response lag of the power module. In practical applications, valid deviations can be confirmed by deviation thresholds, durations, or the number of consecutive samplings. For example, a valid deviation can be determined when the difference between the output power and the currently allocated power exceeds a preset power threshold or exceeds a preset proportion of the currently allocated power, and persists for one or more control cycles. The power threshold, preset proportion, and duration can be set according to the rated power of the charging terminal, the response speed of the power module, the control cycle, and the tolerance range of power fluctuations on site. It is sufficient that the valid power difference affecting shared capacity allocation can be distinguished from normal operational fluctuations; this application does not impose further limitations in this regard.

[0102] It should be noted that the acyclic disturbance absorption graph in this application can be understood as an event-type graph structure temporarily formed for a confirmed power disturbance, based on the original neighborhood communication graph, with unidirectional transmission relationships and no disturbance return path. The original neighborhood communication graph describes the range of state interactions that can occur between charging terminals during normal power distribution. It is usually affected by power distribution access, spatial layout, and communication links, and therefore tends to exhibit local mesh, local ring, or multi-path connection characteristics in actual charging stations. The acyclic disturbance absorption graph is not a replacement for the original neighborhood communication graph, nor does it change the physical communication connections between charging terminals. Instead, during a power disturbance processing, it temporarily determines the allowed transmission direction and acceptance sequence of the disturbance power from the original neighborhood communication graph, so that the same disturbance power can be absorbed step by step along the controlled path.

[0103] The reason for forming an acyclic disturbance absorption graph is that the disturbance power processed in this application has a clear and finite source. This disturbance power originates from the difference between the current allocated power of the disturbance root node and the executable output power constrained by the vehicle BMS, or from the already received power that needs to be rolled back when the disturbance root node resumes output. This power difference only corresponds to one release or one recovery in the shared power capacity and cannot be reused due to the increase in communication paths. If this power change is directly propagated in the original neighborhood communication graph, multiple adjacent nodes may repeatedly receive the same disturbance information through different paths and use it as new acceptable power to participate in local consistency updates, thereby causing the total increase in power of each node to exceed the actual released power difference. This phenomenon is not prominent in single-chain communication relationships, but in three-dimensional fast charging stations, because the same charging area, adjacent floors, the same power distribution cabinet, and local communication gateways may simultaneously form cross-adjacency relationships, multi-path repeated sensing and backflow propagation are more likely to occur.

[0104] For the reasons mentioned above, the essence of the acyclic perturbation absorption graph lies in establishing temporary propagation boundaries and directional constraints for a single finite power perturbation. Through acyclic processing, the same node no longer responds repeatedly to the same perturbation power from multiple sources in a single perturbation event, and the perturbation power is not cyclically transmitted between processed nodes. Thus, the processing logic of perturbation power changes from free diffusion in the original mesh graph to gradual acceptance or transfer along the perturbation hierarchy. Nodes closer to the root node of the perturbation can preferentially absorb perturbation power; only when the current level cannot fully absorb it will the remaining perturbation power continue to enter the next level. This reduces unnecessary synchronization adjustments at distant nodes and allows the transmission range of perturbation power to naturally converge as acceptance is completed.

[0105] It should also be noted that the acyclic nature of the acyclic disturbance absorption graph does not require the original communication network of the charging station to be free of loops, nor does it require the physical disconnection of existing communication links at the charging station. This means that during the current disturbance power processing, the propagation relationship of the disturbance power does not form a closed loop. For multi-path connections that already exist in the original neighborhood communication graph, only directions satisfying the current disturbance processing rules are selected for propagation during disturbance processing; other paths that may lead to backflow or duplicate responses are not included in the calculation of the disturbance power. While preserving the integrity of the original neighborhood communication graph, only temporary constraints are applied to the calculation and propagation process of the disturbance power, balancing the flexibility of normal distributed consistent power allocation with the stability during sudden power deviation processing.

[0106] In one example, the acyclic perturbation absorption map is generated in the following ways:

[0107] S2.1: The disturbance root node is taken as the currently included node, and the charging terminal that has a neighborhood state interaction relationship with the currently included node and has not yet been included in the current power disturbance is determined as the candidate node.

[0108] Specifically, the disturbance root node is the source node of the current power disturbance. For example, if a charging terminal is currently allocated 80kW of power, but can only stably output 50kW due to the constraint of the vehicle's BMS allowing a decrease in charging parameters, then this charging terminal generates 30kW of disturbance power to be absorbed, which serves as the starting point of the acyclic disturbance absorption graph. Although there may be multiple charging terminals adjacent to the disturbance root node in the original neighborhood communication graph, not all adjacent charging terminals will directly participate in the processing of this disturbance power. To prevent this finite disturbance power from freely spreading in the original neighborhood communication graph, it is necessary to determine the range of nodes that the current disturbance event is allowed to pass through, layer by layer, starting from the disturbance root node.

[0109] In this embodiment, candidate nodes can be obtained by reading the adjacency matrix rows corresponding to currently included nodes in the original neighborhood communication graph. Taking the adjacency matrix as an example, if the currently included node is V1, the data row corresponding to V1 in the adjacency matrix is ​​read. When the data in this data row corresponding to nodes V2 and V3 is 1, it can be determined that V2 and V3 have a neighborhood state interaction relationship with V1. Subsequently, it is necessary to determine whether V2 and V3 have been included in the current power disturbance; if they have not been included, they are considered candidate nodes; if they have been included in previous iterations, they will not be considered candidate nodes again. This process can be completed by establishing a temporary inclusion record for the current disturbance event. The temporary inclusion record can include basic data such as the disturbance event identifier, the included node identifier, and the inclusion round. This temporary inclusion record is only used during the current disturbance processing and can be deleted or archived after the disturbance processing ends, without changing the basic connection relationship in the original neighborhood communication graph.

[0110] Furthermore, using "not yet included in the current power disturbance" as a candidate node screening criterion can prevent the same charging terminal from being processed repeatedly in the same disturbance event. In cases where there are loops or multiple paths in the original neighborhood communication graph, the same node may be queried multiple times through different neighboring nodes. For example, node V4 may have neighborhood state interaction relationships with both node V2 and node V3. If V2 and V3 were included in earlier rounds, V4 may be discovered through paths V2 and V3 respectively. By using the already included records, it can be ensured that V4 only undergoes one inclusion judgment in this disturbance event, avoiding the formation of multiple disturbance sources subsequently.

[0111] S2.2: Determine whether the candidate node meets the disturbance absorption entry condition. The disturbance absorption entry condition includes: the charging terminal corresponding to the candidate node has an adjustable power that matches the power disturbance direction, and the adjustable power is greater than or equal to a preset power threshold, wherein the power threshold is determined based on the minimum stable adjustment step size of the power module and the normal power fluctuation range on site.

[0112] Specifically, the adjacency relationships in the original neighborhood communication graph only indicate that two charging terminals can interact with each other, and do not necessarily mean that the charging terminal corresponding to a candidate node is suitable for receiving the current disturbance power. If all adjacent nodes are included in the acyclic disturbance absorption graph based solely on communication relationships, the graph is merely a complete expansion of the original neighborhood communication graph, which cannot distinguish which nodes have actual power regulation capabilities, nor can it reduce the participation of invalid nodes in disturbance propagation.

[0113] In this embodiment, the adjustable power of the candidate node can be determined by the vehicle BMS allowed charging parameters, current output power, rated output capacity, single-cycle adjustment step size, and remaining capacity of the branch of the charging terminal corresponding to the candidate node.

[0114] In one optional implementation, the vehicle-side allowable output upper limit can first be determined based on the allowable charging current, allowable charging voltage, and requested charging power of the vehicle's BMS corresponding to the candidate node. When the vehicle BMS simultaneously feeds back the allowable charging current, allowable charging voltage, and requested charging power, the smaller of the allowable charging current, the power value corresponding to the allowable charging voltage, and the requested charging power can be used as the vehicle-side allowable output upper limit. Subsequently, the vehicle-side allowable output upper limit is compared with the rated output power of the charging terminal corresponding to the candidate node, and the smaller of the two is taken as the allowable output upper limit of the candidate node under the current vehicle state and terminal hardware conditions. If there are still remaining capacity constraints in the power distribution branch or distribution cabinet where the candidate node is located, the output upper limit is further limited to the range allowed by the remaining capacity of the branch, thus obtaining the restricted output upper limit of the candidate node.

[0115] After obtaining the limited output upper limit, the difference between the limited output upper limit and the current output power of the candidate node can be used as the initial increaseable power of the candidate node. Since the power module of the charging terminal is usually not suitable for large step adjustments in a single control cycle, the single-cycle adjustment step size can be set according to the ramp-up capability of the power module or the field control strategy, and the initial increaseable power can be limited within the range of the single-cycle adjustment step size to obtain the adjustable power of the candidate node under the current power disturbance direction. In other words, for the disturbance power to be absorbed formed by the decrease in the output power of the disturbance root node, the adjustable power of the candidate node can be understood as the power margin that can be increased in the current cycle under the joint constraints of the vehicle-side allowance, the terminal rated capacity, the branch capacity, and the single-cycle adjustment step size.

[0116] For example, if the BMS of a candidate node allows a charging current of 160A and a charging voltage of 500V, and the requested charging power is 75kW, then the upper limit of the allowed output on the vehicle side can be determined to be 75kW. If the rated output power of the charging terminal corresponding to the candidate node is 120kW, the current output power is 68kW, the remaining capacity of the branch allows it to increase by a maximum of 10kW, and the single-cycle adjustment step size is 6kW, then the upper limit of the candidate node's output after being constrained by the vehicle side and limited by the terminal's rated capacity will not exceed 75kW, and the initial increase in power relative to the current output power is 7kW. Combining the remaining capacity of the branch of 10kW and the single-cycle adjustment step size of 6kW, the adjustable power of the candidate node in the current cycle can be determined to be 6kW. If the preset power threshold is 3kW, then the candidate node meets the disturbance absorption entry condition; if the final adjustable power is less than 3kW, it will not be included in this acyclic disturbance absorption diagram to avoid nodes with only a small adjustment margin participating in disturbance transmission.

[0117] S2.3: Incorporate the candidate nodes that meet the disturbance absorption entry conditions into the current power disturbance, and determine the candidate nodes incorporated into the current power disturbance as the downstream nodes of the currently incorporated nodes.

[0118] Specifically, once a candidate node meets the disturbance absorption entry conditions, it indicates that the charging terminal corresponding to that candidate node has the actual participation capability under the current disturbance direction, and can absorb part of the disturbance power, or at least participate in subsequent allocation as an actual receiving node in the current absorption level. At this time, it is necessary to change the candidate node from a normal candidate state to an included node in this disturbance event, and establish a one-way transmission relationship from the currently included nodes to the candidate node. The downstream node in this application indicates that in this disturbance event, the disturbance power can be transmitted from the currently included nodes to the candidate node, rather than indicating that there is a fixed control subordination relationship between the charging terminals.

[0119] In this embodiment, each included candidate node can be programmed with its inclusion source, inclusion round, and disturbance event identifier. For example, if the disturbance root node V1 is in round 0, and candidate nodes V2 and V3 are both discovered by V1 and meet the entry conditions, then V2 and V3 can be included in round 1, and the downstream relationships from V1 to V2 and V1 to V3 can be recorded. If subsequent node V4 is discovered by V2 and meets the entry conditions, then V4 is included in round 2, and the downstream relationships from V2 to V4 can be recorded. In cases where the same candidate node is simultaneously discovered by multiple currently included nodes in the same round, its unique inclusion source can be determined according to a preset selection rule. For example, the selection rule could be to select a source node with lower communication latency, a source node with a smaller node number in the current level, a source node with a higher historical acceptance success rate, or a source node located on the same distribution branch as the candidate node. After selecting an inclusion source, the candidate node will no longer respond to inclusion requests from other source paths in this disturbance event.

[0120] Furthermore, determining a unique downstream relationship for candidate nodes is a crucial step for the acyclic disturbance absorption graph to be valid. Taking the nodes mentioned above as an example, the original neighborhood communication graph may simultaneously contain connections like V1-V2, V1-V3, V2-V4, and V3-V4. If V4 receives power from both V2 and V3, the disturbance power may reach V4 via different paths, and might even be repeatedly counted in subsequent consistency updates. By determining a unique source during inclusion, only one disturbance propagation path into V4 can be retained under the current disturbance event; for example, only the path from V2 to V4 can be retained, while the path from V3 to V4 is not used for this disturbance power propagation.

[0121] It should be noted that when the same candidate node corresponds to multiple possible upstream sources within the same iteration round, the specific disturbance propagation path to retain can be determined by referring to a preset path selection rule. This path selection rule does not require selecting a uniquely optimal path based on power allocation effects; it only needs to ensure that the same candidate node retains only one upstream source in the current disturbance event. Since nodes within the same iteration round belong to the same absorption front, their absorption of disturbance power occurs within the same level. The introduction of a candidate node by any already included node in the same level usually does not change the disturbance level to which the candidate node belongs, nor does it change the overall power range that the current absorption front can absorb. Therefore, when multiple upstream sources exist, the retained path can be determined by using a first-come, first-served approach to communication messages, node number order, prioritizing the path with the lowest communication latency, or a pseudo-random selection method based on disturbance event identifiers.

[0122] In one alternative implementation, an event-related pseudo-random path selection rule can be used. Specifically, when a candidate node is discovered simultaneously by multiple currently included nodes, a path selection factor can be generated based on the disturbance event identifier corresponding to the current power disturbance, the candidate node identifier, and the identifiers of each optional upstream node. The optional upstream node whose path selection factor satisfies a preset sorting condition is then determined as the unique upstream node of the candidate node in the current disturbance event.

[0123] The generation method of the path selection factor may include: concatenating the identifiers according to a fixed field order of the disturbance event identifier, candidate node identifier, and optional upstream node identifier to obtain a path selection string for the corresponding optional upstream node; performing a calculation on the path selection string using a preset hash algorithm to obtain a fixed-length hash result, and converting all or a preset number of bits of the hash result into an unsigned integer, or taking the modulus of the converted unsigned integer according to a preset modulus, and determining the resulting value as the path selection factor. The preset hash algorithm may be a deterministic hash algorithm well-known to those skilled in the art, such as a cyclic redundancy check algorithm, an unencrypted hash algorithm, or a message digest algorithm, as long as the same input corresponds to the same path selection factor and the path selection factors corresponding to different optional upstream nodes have sufficient discreteness.

[0124] In one example, the available upstream nodes can be sorted in ascending order of path selection factor, and the available upstream node with the smallest path selection factor is determined as the unique upstream node of the candidate node. The sorting direction is fixed during system initialization and does not change temporarily during the handling of disturbance events within the same charging station. If two or more available upstream nodes have the same path selection factor, they are sorted a second time according to the numerical value corresponding to the available upstream node identifier, from smallest to largest. When the available upstream node identifier is a character identifier, it can be converted to a numerical value according to a preset character encoding and then compared, or it can be compared directly according to dictionary order. Since the node identifier corresponding to each charging terminal is unique, the unique available upstream node can be determined through the above secondary sorting. Therefore, under the same disturbance event, different charging terminals can obtain consistent and reproducible path selection results when calculating based on the same identifier information. When the disturbance event identifier changes, the path selection factors and their sorting relationships of each available upstream node can change accordingly, thereby avoiding the same communication path being fixed in a long-term manner in continuous disturbance events.

[0125] S2.4: Perform an iterative operation, taking the downstream node as the currently included node, and repeat the inclusion operation of the downstream node until there are no more candidate nodes to be included.

[0126] In this embodiment, the downstream nodes added in the previous round can be used to form the set of currently included nodes in the next round, and the adjacency relationships of these nodes in the original neighborhood communication graph can be read respectively. For each currently included node, the charging terminal with which it has a neighborhood state interaction relationship and has not yet been included in the current power disturbance is searched, and it is regarded as a new candidate node. Then, the disturbance absorption entry condition is judged according to the aforementioned condition. If the new candidate node meets the entry condition, it is included in the current power disturbance and recorded as the downstream node of the corresponding currently included node; if it does not meet the entry condition, it is not included in the disturbance absorption graph. For example, V1 is the root node of the disturbance. V2 and V3 are included in the first round. In the second round, V2 and V3 are used as the currently included nodes to continue to query neighboring nodes. If V2 can find V4 and V4 meets the entry condition, V4 is included in the second round. If V3 finds V5 but V5's adjustable power is lower than the preset power threshold, V5 is not included. In the third round, only the nodes newly included in the second round are used to continue to expand until no new candidate node meets the entry condition.

[0127] S2.5: Generate the acyclic perturbation absorption map based on the included nodes and the connection relationships between them.

[0128] In this embodiment, the acyclic disturbance absorption graph can be stored using a directed adjacency list, a directed adjacency matrix, or an event path table. For example, the original neighborhood communication graph contains connections such as V1-V2, V1-V3, V2-V4, V3-V4, and V4-V5. In this disturbance event, V1 is the root node of the disturbance. In the first round, V2 and V3 are included. In the second round, V4 can be discovered by both V2 and V3, but according to the selection rules of the aforementioned embodiment, V2 is determined to be the only source of inclusion for V4. In the third round, V5 is included. The generated acyclic disturbance absorption graph can retain only the directed connections V1 to V2, V1 to V3, V2 to V4, and V4 to V5, while disabling relationships such as V3 to V4 and V4 pointing back to V2 and V4 pointing back to V3, which may cause backflow or duplicate responses. The corresponding data can be recorded as follows: the downstream nodes of V1 include V2 and V3, the downstream nodes of V2 include V4, the downstream nodes of V4 include V5, and V3 and V5 will no longer form backflow connections to the included nodes during this disturbance event.

[0129] Next, we will further elaborate on the absorption of cutting-edge technologies in the method of this application.

[0130] It should be noted that the absorption front in this application can be understood as: in the acyclic disturbance absorption graph, a group of nodes at the same level that actually participate in the acceptance or transfer of disturbance power at the current stage. The acyclic disturbance absorption graph shows the direction and level of disturbance power propagation outward from the disturbance root node. However, in the actual power processing, it is not necessary for all included nodes to act simultaneously, nor is it advisable for nodes far from the disturbance root node to participate in regulation too early. The absorption front is equivalent to forming a power processing boundary on the acyclic disturbance absorption graph that advances outward with the processing progress. The level to which the disturbance power reaches becomes the set of nodes that need to calculate the share of acceptance and perform output regulation at the current stage.

[0131] From the perspective of power balance, the disturbance power is essentially a finite difference between the currently allocated power of the disturbance root node and the executable output power. This finite difference only needs to be accepted or rolled back once, and does not need to be implemented simultaneously in all charging terminals. If all included nodes participate in the acceptance judgment in the same control cycle, although it may speed up the calculation coverage, it will cause more charging terminals to simultaneously perform power increase or decrease actions, easily expanding a local vehicle-side constraint change into a large-scale synchronous adjustment. The absorption front limits the disturbance processing to the current level, giving priority to nodes that are closer to the disturbance root node, have shorter communication paths, and usually have closer power distribution relationships to undertake the adjustment task; only when the adjustable power of the current level is insufficient to absorb the remaining disturbance power will the processing boundary continue to the next level.

[0132] In distributed consensus power allocation scenarios, absorbing fronts can also make the processing of disturbance power clearly phased. Ordinary consensus updates often iterate continuously based on the power difference in the neighborhood, and nodes may repeatedly perceive the same power change in multiple cycles. Absorbing fronts, on the other hand, break down a power disturbance into a sequential process of "processing at the current level and propagating the remaining part," so that the disturbance power to be absorbed in each stage has a clear source: the first absorbing front corresponds to the power difference released or recovered by the root node of the disturbance, and subsequent absorbing fronts correspond to the remaining part that the previous level failed to absorb. As a result, the disturbance power will not spread disorderly between the same level and different levels, and the determination of the subsequent target output power can also correspond to the absorption result of the specific level.

[0133] It should also be noted that the "front" in the absorption front does not represent a physical spatial boundary, nor does it require charging terminals to be arranged at fixed geographical distances. It can be a logical processing boundary formed by the disturbance hierarchy based on the acyclic disturbance absorption diagram. Nodes within the same absorption front can be located in the same power distribution branch, the same charging area, or adjacent communication areas, as long as they belong to the same disturbance hierarchy in the current disturbance event. Multiple nodes within the same front can share the disturbance power to be absorbed, but their respective shares are still constrained by their adjustable power, the vehicle's BMS allowed charging parameters, and the current output state, and therefore will not exceed their own executable range simply because they are within the same front.

[0134] In one example, the absorption front is generated in the following ways:

[0135] Based on the iteration round corresponding to when each included node is included in the current power disturbance, the disturbance level of each included node relative to the disturbance root node is determined, wherein nodes included in the same iteration round belong to the same disturbance level;

[0136] The group of included nodes that have not yet completed disturbance power processing and have the smallest disturbance level is identified as the current absorption frontier. The criterion for whether disturbance power processing has been completed can be the receipt of confirmation of actual received power, which will not be elaborated here.

[0137] Understandably, the perturbation level can be used to represent the logical propagation order of the included nodes from the perturbation root node in this perturbation event, rather than the physical distance or installation level of the charging terminal. Since the acyclic perturbation absorption graph is formed by including nodes outward from the perturbation root node round by round, nodes included in the same round can be regarded as a set of nodes that simultaneously receive perturbation processing tasks of the same level. Determining the group of nodes that has not yet completed perturbation power processing and has the smallest perturbation level as the current absorption front allows the perturbation power to be processed layer by layer in order from near to far.

[0138] For example, if the perturbation root node V1 is at layer 0, and V2 and V3 are included by V1 in the first iteration, then V2 and V3 belong to layer 1; if V4 is included by V2 in the second iteration, then V4 belongs to layer 2. When V2 and V3 have not yet completed the perturbation power absorption, the current absorption front is V2 and V3, while V4 does not participate in the absorption calculation for the time being; only when V2 and V3 cannot fully absorb the current perturbation power to be processed will the unabsorbed perturbation power continue to be transmitted to the layer where V4 is located.

[0139] In yet another example, the method further includes:

[0140] The charging terminal corresponding to the node within the current absorption front edge is controlled to accept the received disturbance power, and when there is unaccepted disturbance power in the charging terminal within the current absorption front edge, the node included in the next disturbance level is updated as a new absorption front edge.

[0141] The update of the absorption front is stopped when the disturbance power is fully absorbed or there are no nodes included in the next disturbance level.

[0142] In one example, the disturbance power is received in the following ways:

[0143] S3.1: Determine the disturbance power to be received corresponding to the current absorption front, wherein the disturbance power to be received corresponding to the first absorption front is determined based on the change in output power of the disturbance root node, and the disturbance power to be received corresponding to subsequent absorption fronts is determined based on the disturbance power that was not received by the previous absorption front.

[0144] Specifically, the disturbance power to be absorbed is the finite power difference that the current absorption front needs to handle, rather than the total power that each charging terminal within the current absorption front can request to increase on its own. For the first absorption front, the disturbance power to be absorbed comes from the amount of output power change that the disturbance root node cannot continue to execute due to changes in the vehicle's BMS-permitted charging parameters.

[0145] In this embodiment, the change in output power of the disturbance root node can be determined based on the current allocated power of the disturbance root node, the executable output power after vehicle-side constraints, and the actual output power following state. The current allocated power can be the target output power determined in the power allocation cycle before the disturbance occurs, and the executable output power can be determined based on the vehicle BMS's allowed charging current, allowed charging voltage, requested charging power, terminal rated output capacity, and local limiting conditions. To avoid mistaking short-term sampling jitter for disturbance power, a confirmation condition can be set when determining the change in output power. For example, if the difference between the current allocated power and the executable output power is greater than 3kW and remains for two consecutive control cycles, it can then be confirmed as valid disturbance power. Taking an 80kW current allocated power as an example, if the executable output power after vehicle-side constraints is 78.5kW, the difference is only 1.5kW, which can be treated as normal operating fluctuation; if the executable output power after vehicle-side constraints is 68kW, the difference is 12kW, and it has not recovered for two consecutive control cycles, then 12kW can be confirmed as the source of power to be received in this disturbance event.

[0146] S3.2: Determine the adjustable power of each charging terminal within the current absorption front, and determine the disturbance power acceptance share of each charging terminal according to the adjustable power of each charging terminal and the disturbance power to be accepted, wherein the sum of the disturbance power acceptance shares is less than or equal to the disturbance power to be accepted, and the disturbance power acceptance share of any charging terminal is less than or equal to its corresponding adjustable power.

[0147] Specifically, although the charging terminals within the current absorption front are at the same disturbance level, their vehicle states, output states, power module margins, and power distribution branch constraints are not the same, making it impossible to simply distribute the disturbance power to be absorbed equally. If a charging terminal's vehicle BMS is close to current-limiting, or its current output is close to its rated power, it is not suitable to allocate a large share of the disturbance power, even if it is located at the current absorption front. Therefore, it is necessary to calculate the adjustable power of each charging terminal within the current absorption front, and then determine the disturbance power absorption share for each charging terminal by using the disturbance power to be absorbed as the total upper limit and the adjustable power of each terminal as the individual upper limit. The absorption share is not the expected power of the charging terminal, nor is it the neighborhood consistency difference itself, but rather the amount of disturbance power that the charging terminal is allowed to actually increase or release during this disturbance event.

[0148] It is important to emphasize that the aforementioned total constraint on the share of power allocation can only ensure that the disturbance power allocated at the current absorption front does not exceed the disturbance power to be allocated. However, when there are multiple charging terminals that can be allocated within the same absorption front, the question of how to determine the allocation order and basis remains. If allocation is always based on node number, communication message arrival order, maximum adjustable power priority, or a fixed ratio, although the implementation method is relatively straightforward, in fast charging scenarios with continuous vehicle access, frequent current limiting by vehicle BMS, and multiple disturbance events occurring in succession, the same batch of charging terminals may be repeatedly placed in the priority position. After multiple power allocation cycles, some local distribution branches may bear more power increase operations for a long time, and the power modules of some charging terminals may also frequently perform ramp-up adjustments, thus gradually manifesting the originally local disturbance power handling as branch load bias and local power concentration.

[0149] On the other hand, the adjustable power reported or calculated by each charging terminal within the current absorption front is not always completely consistent with the power it can actually receive subsequently. Vehicle BMS messages may be delayed, the battery-side allowable current may be reduced again within a short period, and the local power module of the charging terminal may fail to achieve the previously declared adjustable power due to temperature, current limiting, or protection strategies. If the aforementioned adjustable power is directly used as the basis for proportional allocation, nodes with a larger declared adjustable power will receive a larger share of the power. After their actual reception fails, the remaining disturbance power needs to be passed to the next absorption front, resulting in the same disturbance power undergoing multiple allocations, recovery, and redistribution. Although this process can still be completed through subsequent level compensation, it increases the number of power adjustments and makes the absorption front advancement process less smooth.

[0150] Based on the above, this application does not simply pursue "the higher the adjustable power, the higher the priority" in determining the share of acceptance. Instead, it takes into account both the reliability of the adjustable power and the dispersion of the acceptance priority. The source of this idea can be understood in two aspects: First, it draws on the approach of fault-tolerant consistency control that does not completely depend on the state declaration of a single node. That is, the adjustable power declared by the charging terminal needs to be verified with the charging parameters allowed by the vehicle BMS, the current output power, the historical acceptance results, and the actual output following situation to avoid abnormal declarations directly affecting the disturbance power allocation results. Second, it draws on the approach of perturbation processing of fixed priorities in load balancing scheduling. That is, after confirming that a node has the effective acceptance conditions, the acceptance order is no longer fixed in the long term, but the priority acceptance node changes appropriately under different disturbance events.

[0151] This allows for a "constraint-first, disturbance-later" approach to node assignment. Specifically, before assignment begins, nodes lacking effective adjustable power, whose adjustable power declarations are inconsistent with vehicle-side constraints, or those with a history of recent assignment failures are weakened or eliminated from the primary assignment criteria. This ensures that nodes entering the assignment range at least meet basic executable conditions. Within the candidate node pool that has passed reliability assessment, node number, adjustable power size, or communication order are no longer fixed as the sole ranking criteria. Instead, assignment priority is correlated with the current disturbance event, thus creating a variable assignment order across different disturbance events.

[0152] This approach does not randomly allocate disturbance power to any charging terminal, nor does it reduce the controllability of the power allocation process. Changes in the acceptance order occur within the range of nodes that have already met the acceptance criteria, and the acceptance share remains limited by the total amount of disturbance power to be accepted and the reliable adjustable power of each individual charging terminal. In other words, randomness or disturbance is only used to break up the long-term fixed priorities among qualified candidate nodes, and not to overcome vehicle BMS constraints, terminal rated capacity, or branch capacity constraints.

[0153] Based on the above ideas, the current power assignment of disturbances within the absorption front is no longer simply a matter of allocating a finite power difference. Instead, the assignment criteria are reliably corrected before allocation, and the assignment priority is adjusted within a reliable range. This reduces the impact of abnormal adjustable power claims on the determination of assignment shares and also reduces the long-term concentration of assignment nodes caused by fixed allocation rules, allowing the power regulation burden under continuous disturbance events to be distributed more evenly among nodes with assignment capabilities. For multi-terminal fast charging scenarios with shared distribution capacity, this approach helps reduce repeated ramping in local branches, frequent assignment by local terminals, and repeated compensation in the subsequent absorption front, enabling the disturbance power to be absorbed more reliably within the current absorption front.

[0154] In one example, the method for determining the disturbance power share includes:

[0155] Based on the adjustable power of each charging terminal within the current absorption frontier, the vehicle BMS allowed charging parameters, and historical acceptance results, the adjustable power of each charging terminal is reliably verified to obtain the reliable adjustable power corresponding to each charging terminal.

[0156] Charging terminals with reliable adjustable power greater than or equal to the preset acceptance lower limit are identified as candidate acceptance nodes.

[0157] Based on the disturbance event identifier, disturbance level, disturbance root node identifier, and node identifier of each candidate node corresponding to the current power disturbance, a succession ranking factor is generated for each candidate node, and the succession order of each candidate node is determined according to the succession ranking factor.

[0158] According to the acceptance order, the disturbance power acceptance share of each candidate node is determined sequentially within the range of the disturbance power to be accepted, wherein the disturbance power acceptance share of any candidate node is less than or equal to its corresponding reliable adjustable power, and the sum of the disturbance power acceptance shares is less than or equal to the disturbance power to be accepted.

[0159] In practical implementation, the reliable verification of adjustable power can be completed before the candidate node enters the acceptance allocation. For any charging terminal within the current absorption front, a theoretical adjustable upper limit, jointly limited by vehicle-side constraints and terminal hardware constraints, can be obtained based on the vehicle BMS's allowed charging current, allowed charging voltage, requested charging power, current output power, rated output capacity, and single-cycle adjustment step size. Subsequently, the adjustable power reported by the charging terminal or calculated locally is compared with the theoretical adjustable upper limit, and the portion exceeding the theoretical adjustable upper limit is not used as a reliable acceptance basis. For example, if a charging terminal reports an adjustable power of 18kW, but after conversion based on the vehicle BMS's allowed charging current and allowed charging voltage, its vehicle side can only allow a maximum increase of 12kW, and the single-cycle adjustment step size is 8kW, then the reliable verification for this cycle will not use 18kW as the allocation basis, but will truncate it to no more than 8kW. If the remaining capacity of the branch where the terminal is located only allows an increase of 6kW, then the reliable adjustable power is further limited to 6kW. This step-by-step truncation avoids situations where excessively large claims caused by delayed BMS messages, outdated local caches, or abnormal power module states are directly included in the share calculation.

[0160] Trustworthy verification can also be corrected by combining historical connection results. Historical connection results can record the assigned connection share, actual output increase, unaccepted amount, and connection completion rate of the same charging terminal in several recent disturbance events. For example, the most recent 5 or 10 disturbance events can be used as a statistical window; if a charging terminal's actual connection power is lower than 70% of its assigned connection share in 3 out of the most recent 5 events, where 3 events and the 70% threshold are determined based on historical normal connection deviation statistics, then its recent connection stability can be considered insufficient, and its adjustable power can be reduced. For example, if a terminal currently calculates an adjustable power of 10kW based on vehicle-side constraints and hardware constraints, but the median value of the actual stable connection power in the most recent 5 events is 5kW, and there are two consecutive records of insufficient connection, then the trusted adjustable power can be corrected to no more than 5kW to 6kW, or it can be excluded from the priority connection node in the current disturbance event. The preset lower limit can be determined based on the minimum stable adjustment step size of the power module. For example, the minimum stable adjustment step size for a 120kW DC fast charging terminal is 3kW, and the tolerance range for ordinary power fluctuations in the field is about 2kW, so the preset lower limit can be set to 3kW. If the rated power is high and a single adjustment that is too small is easily swallowed up by closed-loop fluctuations, 3% to 5% of the rated power can also be used as a reference, for example, 5kW to 9kW for a 180kW terminal. Terminals with reliable adjustable power below this lower limit can be temporarily excluded from the set of candidate nodes to reduce ineffective adjustments caused by small acceptance shares.

[0161] After determining the candidate nodes for acceptance, the acceptance order should not be fixed in the order of node numbers or adjustable power values ​​for a long period. Optionally, the disturbance event identifier, disturbance level, disturbance root node identifier, and candidate node identifier can be combined into an event-related string, and the acceptance ranking factor can be obtained through cyclic redundancy check, hash operation, or pseudo-random number seed generation. Then, the acceptance order within the current disturbance event is determined according to the size of the acceptance ranking factor. For example, if the current disturbance event identifier is E001, the disturbance level is 1, the disturbance root node is V1, and the candidate nodes for acceptance include V2, V3, and V4, then three ranking factors can be generated based on "E001-1-V1-V2", "E001-1-V1-V3", and "E001-1-V1-V4" respectively, and they can be allocated in ascending order of ranking factor. This ranking is reproducible within the same disturbance event and will not cause order conflicts due to repeated calculations by different charging terminals. In the next disturbance event, the acceptance order will also change due to changes in the disturbance event identifier or the disturbance root node. During share allocation, the power to be accepted is deducted sequentially from the disturbance power to be accepted in the acceptance order. Each candidate node can obtain a share within its trusted adjustable power range at most, until the disturbance power to be accepted is fully allocated or all candidate nodes reach the trusted adjustable power limit. For example, if the disturbance power to be accepted is 14kW, and the trusted adjustable powers corresponding to V3, V2, and V4 after ranking are 6kW, 5kW, and 7kW respectively, then V3 can be allocated 6kW, V2 5kW, and V4 the remaining 3kW in sequence; the total share is 14kW, which does not exceed the disturbance power to be accepted, and each share does not exceed the corresponding trusted adjustable power. Thus, the acceptance allocation retains the finite power total boundary, avoids fixed nodes repeatedly accepting priority in consecutive disturbance events, and weakens the impact of abnormal adjustable power declarations on the allocation results.

[0162] The disturbance event identifier can be generated by combining the disturbance root node identifier, the disturbance trigger time, and the event sequence number according to a preset field order. The disturbance trigger time can be the system time, control cycle number, or sampling cycle number corresponding to the first time the disturbance root node confirms a valid deviation of the output power relative to the currently allocated power; the event sequence number can be maintained by the disturbance root node and incremented each time a new power disturbance is confirmed. When the same disturbance root node triggers multiple disturbance events consecutively within the same time granularity, each disturbance event can be distinguished by different event sequence numbers.

[0163] In practical implementation, the disturbance event identifier can be generated by the disturbance root node and transmitted to downstream nodes along with the disturbance power information. This ensures that all charging terminals participating in the same disturbance handling process use the same disturbance event identifier, without generating new event identifiers separately. Since different disturbance root nodes have different node identifiers, the same disturbance root node corresponds to different disturbance trigger times at different trigger times, and multiple disturbance events within the same trigger time correspond to different event sequence numbers. Therefore, the disturbance event identifier is unique within the current charging station's disturbance event record range. When it is necessary to distinguish disturbance events across charging stations, a charging station identifier or a station-level control device identifier can be added to the disturbance event identifier. This avoids confusion in the acceptance records, power shares, and fallback paths of different disturbance events and provides a unified input for subsequently generating reproducible and event-dependent acceptance sorting factors.

[0164] S3.3: Control each charging terminal to adjust its output power according to the corresponding disturbance power acceptance share, and determine the unaccepted disturbance power to be transmitted to the next absorption front based on the difference between the disturbance power to be accepted and the actual power accepted by each charging terminal.

[0165] In yet another example, embodiments of this application provide a power allocation system based on a distributed consensus algorithm, the system comprising:

[0166] The data acquisition module is used to acquire the neighborhood communication relationship between multiple charging terminals in the charging station, generate the original neighborhood communication map, and acquire the output power of each charging terminal and the vehicle BMS allowed charging parameters.

[0167] The disturbance absorption graph generation module is used to determine the corresponding charging terminal as the disturbance root node when the vehicle BMS of any charging terminal allows changes in charging parameters, and to determine the downstream node based on the disturbance root node in the original neighborhood communication graph to generate an acyclic disturbance absorption graph.

[0168] The disturbance power receiving module is used to generate an absorption front according to the disturbance level in the acyclic disturbance absorption diagram, so that the charging terminal corresponding to the node located in the absorption front receives the disturbance power based on its own adjustable power, and transfers the unreceived disturbance power to the next level charging terminal.

[0169] The target power determination module is used to determine the corresponding target output power based on the absorption result of the disturbance power by each charging terminal, and to back off the absorbed disturbance power in the opposite direction of the acyclic disturbance absorption diagram when the output power of the disturbance root node recovers.

[0170] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A power allocation method based on a distributed consensus algorithm, characterized in that, The method includes: Obtain the neighborhood communication relationship between multiple charging terminals within the charging station, generate the original neighborhood communication graph, and obtain the output power of each charging terminal and the vehicle BMS allowed charging parameters; When the vehicle BMS of any charging terminal changes its allowed charging parameters, causing its output power to deviate from the currently allocated power, the corresponding charging terminal is identified as the disturbance root node, and the downstream node is determined based on the disturbance root node in the original neighborhood communication graph to generate an acyclic disturbance absorption graph. In the acyclic disturbance absorption diagram, an absorption front is generated according to the disturbance level, so that the charging terminal corresponding to the node located in the absorption front can accept the disturbance power based on its own adjustable power, and transfer the unaccepted disturbance power to the next level charging terminal. The target output power is determined based on the absorption result of the disturbance power by each charging terminal, and when the output power of the disturbance root node is restored, the absorbed disturbance power is retracted in the opposite direction of the acyclic disturbance absorption diagram.

2. The method according to claim 1, characterized in that, The neighborhood communication relationship is used to characterize whether there is a neighborhood state interaction relationship between any two charging terminals. The methods for obtaining the neighborhood communication relationship include: By acquiring at least one of the power distribution access identifier, installation location identifier, and communication connection status of each charging terminal, two charging terminals that meet the preset adjacent conditions are identified as having a neighborhood status interaction relationship. The preset adjacent conditions include at least one of the following: connected to the same power distribution branch, connected to the same power distribution cabinet, located in the same charging area, located in adjacent charging areas, or having a direct communication link.

3. The method according to claim 2, characterized in that, The original neighborhood communication graph is constructed in the following ways: Each charging terminal is used as a graph node. Based on the neighborhood communication relationship, a connection edge is established between two graph nodes with a neighborhood state interaction relationship. An adjacency matrix is ​​generated based on each graph node and its corresponding connection edge. The original neighborhood communication graph is generated based on the adjacency matrix, wherein the original neighborhood communication graph is used to characterize the node connection relationship that allows each charging terminal to directly interact with the state during the power allocation process.

4. The method according to claim 1, characterized in that, The vehicle BMS allowed charging parameters are used to characterize the charging constraints that the vehicle battery management system allows the charging terminal to output in the current charging state. The vehicle BMS allowed charging parameters include allowed charging current, allowed charging voltage and requested charging power. The vehicle BMS allows charging parameters to be obtained through the charging communication link between the charging terminal and the vehicle battery management system. The charging communication link includes at least one of the following: the vehicle communication interface after the charging gun is connected, the charging controller communication interface, or the charging protocol message exchange link.

5. The method according to claim 1, characterized in that, The generation method of the acyclic perturbation absorption map includes: The disturbance root node is taken as the currently included node, and the charging terminal that has a neighborhood state interaction relationship with the currently included node and has not yet been included in the current power disturbance is determined as the candidate node. Determine whether the candidate node meets the disturbance absorption entry condition. The disturbance absorption entry condition includes: the charging terminal corresponding to the candidate node has an adjustable power that matches the power disturbance direction, and the adjustable power is greater than or equal to a preset power threshold. Candidate nodes that meet the disturbance absorption entry conditions are included in the current power disturbance, and the candidate nodes included in the current power disturbance are determined as the downstream nodes of the currently included nodes; Perform iterative operations, taking downstream nodes as currently included nodes, and repeat the inclusion operation for downstream nodes until there are no more candidate nodes to be included. The acyclic perturbation absorption map is generated based on the included nodes and their connection relationships.

6. The method according to claim 5, characterized in that, The absorption front is generated in the following ways: Based on the iteration round corresponding to when each included node is included in the current power disturbance, the disturbance level of each included node relative to the disturbance root node is determined, wherein nodes included in the same iteration round belong to the same disturbance level; The group of included nodes that has not yet completed perturbation power processing and has the smallest perturbation level is identified as the current absorption frontier.

7. The method according to claim 6, characterized in that, The method further includes: The charging terminal corresponding to the node within the current absorption front is controlled to accept the received disturbance power, and when there is unaccepted disturbance power in the charging terminal within the current absorption front, the node included in the next disturbance level is updated as a new absorption front. The update of the absorption front is stopped when the disturbance power is fully absorbed or there are no nodes included in the next disturbance level.

8. The method according to claim 1, characterized in that, The methods for receiving the disturbance power include: Determine the disturbance power to be received corresponding to the current absorption front. The disturbance power to be received corresponding to the first absorption front is determined based on the change in output power of the disturbance root node, and the disturbance power to be received corresponding to subsequent absorption fronts is determined based on the disturbance power that was not received by the previous absorption front. The adjustable power of each charging terminal within the current absorption front is determined respectively, and the disturbance power acceptance share of each charging terminal is determined according to the adjustable power of each charging terminal and the disturbance power to be accepted. The sum of the disturbance power acceptance shares is less than or equal to the disturbance power to be accepted, and the disturbance power acceptance share of any charging terminal is less than or equal to its corresponding adjustable power. The output power of each charging terminal is adjusted according to the corresponding disturbance power acceptance share, and the unaccepted disturbance power to be transmitted to the next absorption front is determined based on the difference between the disturbance power to be accepted and the actual power accepted by each charging terminal.

9. The method according to claim 8, characterized in that, The method for determining the share of disturbance power absorption includes: Based on the adjustable power of each charging terminal within the current absorption frontier, the vehicle BMS allowed charging parameters, and historical acceptance results, the adjustable power of each charging terminal is reliably verified to obtain the reliable adjustable power corresponding to each charging terminal. Charging terminals with reliable adjustable power greater than or equal to the preset acceptance lower limit are identified as candidate acceptance nodes. Based on the disturbance event identifier, disturbance level, disturbance root node identifier, and node identifier of each candidate node corresponding to the current power disturbance, a succession ranking factor is generated for each candidate node, and the succession order of each candidate node is determined according to the succession ranking factor. According to the acceptance order, the disturbance power acceptance share of each candidate node is determined sequentially within the range of the disturbance power to be accepted, wherein the disturbance power acceptance share of any candidate node is less than or equal to its corresponding reliable adjustable power, and the sum of the disturbance power acceptance shares is less than or equal to the disturbance power to be accepted.

10. A power allocation system based on a distributed consensus algorithm, implemented based on the power allocation method based on a distributed consensus algorithm as described in any one of claims 1-9, characterized in that, The system includes: The data acquisition module is used to acquire the neighborhood communication relationship between multiple charging terminals in the charging station, generate the original neighborhood communication map, and acquire the output power of each charging terminal and the vehicle BMS allowed charging parameters. The disturbance absorption graph generation module is used to determine the corresponding charging terminal as the disturbance root node when the vehicle BMS of any charging terminal allows changes in charging parameters, and to determine the downstream node based on the disturbance root node in the original neighborhood communication graph to generate an acyclic disturbance absorption graph. The disturbance power receiving module is used to generate an absorption front according to the disturbance level in the acyclic disturbance absorption diagram, so that the charging terminal corresponding to the node located in the absorption front receives the disturbance power based on its own adjustable power, and transfers the unreceived disturbance power to the next level charging terminal. The target power determination module is used to determine the corresponding target output power based on the absorption result of the disturbance power by each charging terminal, and to back off the absorbed disturbance power in the opposite direction of the acyclic disturbance absorption diagram when the output power of the disturbance root node recovers.