Mobile energy storage vehicle cluster power grid support scheduling method

Abstract

The present application belongs to the technical field of energy storage scheduling, and particularly relates to a mobile energy storage vehicle cluster power grid support scheduling method, comprising: calculating the upper limit of adjustable output of each vehicle under the condition of not triggering the derating protection, and summarizing to form an upper limit of adjustable output vector; calculating the maximum safe power of the cluster based on the upper limit of adjustable output vector, distributing and executing the initial power instruction for each vehicle; during the response event execution, calculating the power gap corresponding to the derated vehicle, generating the updated power instruction and re-distributing; after the response event ends, extracting the measured electrical data of each vehicle for online internal resistance identification to update the upper limit of adjustable output, and determining the low availability state vehicle based on the distribution characteristics of the updated upper limit of adjustable output and eliminating it. The present application effectively reduces the occurrence rate of the derating event during the response event execution, and improves the stability of the cluster maintaining the declared power.

Description

Technical Field

[0001] This invention relates to the field of energy storage dispatching technology, and more specifically, to a method for grid support dispatching of mobile energy storage vehicle clusters. Background Technology

[0002] Against the backdrop of new power system construction, distributed energy storage resources, carried by mobile energy storage vehicles, are participating in the grid ancillary services and demand response market in batches through load aggregator platforms. This has become an important way for energy storage assets along highways to realize their economic value. Mobile energy storage vehicle clusters have the advantages of flexible geographical location and clear response boundaries, demonstrating flexibility that fixed energy storage cannot replicate in ancillary service scenarios such as peak shaving and valley filling, and near real-time response. The settlement rules for participating in near real-time response stipulate that within 15 minutes of the start of the execution period, the actual voltage drop power of the cluster must not be less than 50% of the declared power; otherwise, the response will be deemed invalid, and all standby capacity subsidies for the month will be cancelled.

[0003] Existing mobile energy storage vehicle cluster scheduling methods typically employ two approaches: one is to issue the same power command to all vehicles within the cluster, i.e., even distribution scheduling; the other is to allocate target power to each vehicle proportionally based on the battery stack state of charge (SBC) of each vehicle. Both methods assume that the actual discharge capacity of each vehicle is only constrained by the SBC. Therefore, vehicles with a higher SBC receive greater discharge power, while vehicles with a lower SBC receive less discharge power. This logic is intuitive and easy to implement in engineering. The platform receives telemetry data such as battery stack SBC, voltage, current, and temperature reported by each vehicle's BMS at 1-second intervals via the MQTT protocol, as well as AC / DC side power and alarm status reported by the PCS, ensuring a complete data link.

[0004] However, as the cumulative number of charge-discharge cycles for some vehicles in the fleet increased, the internal resistance of individual battery cells continued to rise, and their health status declined. When a near-real-time response event was initiated, the dispatch platform issued power commands to each vehicle according to the battery stack's state of charge (SOC). Vehicles with deteriorated SOC were required to output a larger power output matching their SOC. Under high-current discharge conditions, the terminal voltage of high-resistance batteries dropped rapidly. When the DC bus voltage fell below the PCS's DC bus undervoltage protection setting, the PCS triggered derating protection, and the vehicle's actual output power was automatically limited below the rated value. Because near-real-time response requires power voltage reduction to be completed within 15 minutes, and PCS derating events occurring in vehicles with deteriorated SOC often occur concentratedly within minutes of response initiation, the cluster's actual total output power could not maintain the declared power during the critical window period. The entire cluster's response was deemed invalid, and all standby capacity subsidies for the month were cancelled. This technical solution aims to address the problem that during the near real-time grid support event execution process of mobile energy storage vehicle clusters, the voltage drop at the high internal resistance vehicle terminals triggers PCS derating protection, resulting in the actual output power of the cluster being lower than the effective response threshold corresponding to the declared power, thus causing invalid responses for the entire batch. Summary of the Invention

[0005] To address the technical problem of invalid responses across the entire batch due to voltage drops at the high-resistance vehicle terminals triggering derating protection during near-real-time grid support event execution in mobile energy storage vehicle clusters, this invention provides a grid support scheduling method for mobile energy storage vehicle clusters.

[0006] This invention provides a grid-supported scheduling method for mobile energy storage vehicle clusters, comprising: S1: acquiring real-time parameters of the battery management system of each vehicle in the mobile energy storage vehicle cluster, combining them with a preset DC bus undervoltage protection setting, calculating the adjustable output limit of each vehicle under the condition of not triggering derating protection, and summarizing them to form an adjustable output limit vector; S2: calculating the maximum safe power of the cluster based on the adjustable output limit vector, combining the total target power issued by the grid side and the margin coefficient determined based on the historical derating rate, allocating initial power commands to each vehicle and issuing them for execution; S3: during the execution of the response event, monitoring the alarm status of each vehicle to identify derating vehicles, calculating the power gap corresponding to the derating vehicles, and distributing the power gap based on the available margin of normally operating vehicles, generating updated power commands and reissuing them; S4: after the response event ends, extracting the measured electrical data of each vehicle for online internal resistance identification to update the adjustable output limit, and determining low-availability vehicles based on the updated adjustable output limit distribution characteristics, and removing the low-availability vehicles from the resource application list in the next scheduling cycle.

[0007] By adopting the above technical solution, this invention introduces the critical undervoltage derivation of the battery equivalent circuit model into the power allocation stage before the event, so that the command issued by each vehicle is strictly limited to the current safe discharge boundary, eliminating the condition that the voltage drop at the terminal of a high-resistance vehicle triggers undervoltage derating under high current load, and effectively reducing the occurrence rate of derating events within the cluster during the execution of response events.

[0008] Preferably, the step of acquiring real-time parameters of the battery management system of each vehicle in the mobile energy storage vehicle cluster, combining them with a preset DC bus undervoltage protection setting, calculating the adjustable output limit of each vehicle under the condition of not triggering derating protection, and summarizing them to form an adjustable output limit vector includes: extracting the current state of charge of the battery stack and the average internal resistance of the battery stack for each vehicle; querying the open-circuit voltage characteristic curve based on the current state of charge of the battery stack to obtain the corresponding open-circuit voltage of the battery stack; calculating the difference between the open-circuit voltage of the battery stack and the DC bus undervoltage protection setting, and dividing the difference by the average internal resistance of the battery stack to obtain the critical current; multiplying the critical current by the DC bus undervoltage protection setting and multiplying by a preset power AC-DC conversion equipment inverter efficiency to obtain the adjustable output limit of each vehicle converted to the AC side under the condition of not triggering derating protection, and summarizing them to form an adjustable output limit vector.

[0009] By adopting the above technical solution, the present invention combines the real-time reported internal resistance data with the calibration curve and directly embeds the critical undervoltage derivation into the output upper limit calculation process, so that the safe discharge boundary of each vehicle can be accurately quantified before the event starts, providing real-time constraint input with internal resistance as the core dimension for subsequent power distribution.

[0010] Preferably, the step of calculating the maximum safe power of the cluster based on the adjustable output upper limit vector, combining the total target power issued by the grid side and the margin coefficient determined based on the historical derating rate, and allocating and issuing initial power commands to each vehicle includes: summing the adjustable output upper limits of all vehicles in the adjustable output upper limit vector to obtain the maximum safe power of the cluster; statistically analyzing the proportion of vehicles that experienced derating in historical response events as the historical derating rate, and determining the margin coefficient based on the historical derating rate and a preset lower limit value of the margin coefficient; calculating the ratio of the adjustable output upper limit of each vehicle to the maximum safe power; if the maximum safe power is not less than the total target power, multiplying the ratio, the total target power, and the margin coefficient together; if the maximum safe power is less than the total target power, replacing the total target power with the maximum safe power for multiplication, and finally obtaining the initial power command corresponding to each vehicle and issuing it for execution.

[0011] Preferably, the monitoring of the alarm status of each vehicle to identify the derating vehicle includes: continuously receiving the real-time alarm status of the AC / DC power conversion equipment contained in each vehicle; detecting whether a DC bus undervoltage alarm indicator appears in the real-time alarm status; if the DC bus undervoltage alarm indicator appears, simultaneously checking whether the AC side power derating alarm indicator is set; if the AC side power derating alarm indicator is set simultaneously, the corresponding vehicle is finally identified as the derating vehicle.

[0012] By adopting the above technical solution, this invention introduces a dual status word verification mechanism for DC bus undervoltage and AC side power derating in the vehicle derating identification process. This mechanism requires that the physical cause of the derating and the actual derating result be set synchronously, effectively eliminating single voltage false alarm interference caused by power grid transient disturbances, bus voltage spikes, or communication measurement noise, completely cutting off the false triggering path of the power transfer compensation procedure, and significantly improving the anti-interference capability and execution accuracy of the cluster dispatch control system under complex operating conditions.

[0013] Preferably, the step of calculating the power gap corresponding to the derated vehicle, allocating the power gap based on the available margin of normally operating vehicles, generating an updated power command, and reissuing it includes: reading the measured power of the derated vehicle, calculating the difference between the current executed power command of the derated vehicle and the measured power to obtain the power gap; calculating the difference between the adjustable output limit of the normally operating vehicle and the corresponding current executed power command to obtain the available margin of the normally operating vehicle; calculating the proportion of the available margin of each normally operating vehicle in the total available margin of all normally operating vehicles; weighting the power gap based on the proportion and adding it to the corresponding current executed power command; if the sum is greater than the corresponding adjustable output limit, limiting it to the adjustable output limit, and finally generating the updated power command and reissuing it.

[0014] By adopting the above technical solution, the present invention uses weighted power transfer calculation based on reserved margin to ensure that the actual total output of the cluster remains above the effective response threshold corresponding to the declared power after the devaluation event occurs, thus meeting the conditions for valid recognition of near real-time response and avoiding the risk that the devaluation of a single vehicle will drag down the effectiveness of the entire cluster response.

[0015] Preferably, the step of extracting measured electrical data of each vehicle for online internal resistance identification to update the adjustable output limit includes: reading the terminal voltage sampling sequence and discharge current sampling sequence of each vehicle during the execution of the response event; extracting the time period from the discharge current sampling sequence where the relative change amplitude of the discharge current is lower than the steady-state determination threshold as a steady-state discharge segment; within the steady-state discharge segment, obtaining the state of charge of the battery stack corresponding to the steady-state discharge segment to query the open-circuit voltage at this time; calculating the updated average internal resistance of the battery stack based on the steady-state mean of the open-circuit voltage, the terminal voltage sampling sequence, and the discharge current sampling sequence according to Ohm's law formula; recalculating based on the updated average internal resistance of the battery stack to finally obtain the updated adjustable output limit.

[0016] By adopting the above technical solution, the present invention directly uses the measured electrical data generated by the event itself to identify the working condition online, without the need for special internal resistance measurement excitation, so that the upper limit vector of output can be refreshed in time after each event, and continuously track the dynamic changes of the fleet health status.

[0017] Preferably, the step of extracting the time period in which the relative change amplitude of the discharge current is lower than the steady-state determination threshold from the discharge current sampling sequence as the steady-state discharge segment includes: obtaining a historical discharge current sequence from historical events of the same vehicle model; calculating a preset percentile of the absolute value of the current change between adjacent sampling points of the historical discharge current sequence; calculating the average discharge current of the historical discharge current sequence; and dividing the preset percentile by the average discharge current to finally obtain the steady-state determination threshold to extract the steady-state discharge segment.

[0018] Preferably, the step of determining low-availability vehicles based on the updated adjustable output limit distribution characteristics and removing them from the resource application list in the next scheduling cycle includes: obtaining the updated adjustable output limit of all registered vehicles in the cluster and calculating the median; extracting the minimum value of the adjustable output limit of normally operating vehicles in the cluster that have not experienced derating protection in historical response events and calculating the ratio of the minimum value to the median at the time of the historical response event; calculating the rolling average of the ratios from multiple historical response events to obtain the lower bound ratio of normally operating vehicle output; multiplying the median by the lower bound ratio of normally operating vehicle output to obtain the removal threshold; and finally determining vehicles whose updated adjustable output limit is lower than the removal threshold as low-availability vehicles and removing them.

[0019] Preferably, after calculating the maximum safe power of the cluster based on the adjustable output upper limit vector, the method further includes: comparing the maximum safe power with the total target power; if the maximum safe power is less than the total target power, generating and sending an early warning notification of insufficient response capability; using the maximum safe power as the internal execution target for this event, and attaching the maximum safe power to the settlement declaration materials after the event ends.

[0020] Preferably, the step of acquiring the real-time parameters of the battery management system of each vehicle in the mobile energy storage vehicle cluster includes: establishing a remote communication connection with the battery management system of each vehicle; continuously receiving telemetry data frames reported by each vehicle according to a preset acquisition cycle; parsing the telemetry data frames, and finally extracting the real-time parameters of the battery management system of each vehicle.

[0021] The technical solution of the present invention has the following beneficial technical effects: This invention transforms the traditional static state of charge allocation into a real-time constraint allocation based on dynamic internal resistance by introducing the critical undervoltage derivation of the battery equivalent circuit model. This avoids the risk of high internal resistance vehicles triggering derating protection due to voltage drop under high current load, significantly improves the stability of the cluster in maintaining total output power within the near real-time response window, and ensures the execution quality of grid ancillary services.

[0022] Furthermore, this invention combines weighted power transfer calculation with reserved margin with online internal resistance identification after an event. This not only enables rapid compensation of power gaps during sudden derating but also achieves continuous dynamic tracking of fleet health. The low-availability state determination criteria derived from historical data can adaptively remove vehicles in low-availability states, further optimizing the overall health of the resource application list and effectively maintaining the long-term reliability and economic benefits of the cluster's participation in power grid support dispatching. Attached Figure Description

[0023] Figure 1 This is a flowchart illustrating a mobile energy storage vehicle cluster grid support scheduling method according to the present invention; Figure 2 This is a schematic diagram illustrating the line graph comparing the deviation of the cluster's actual output in each month of the year in an embodiment of the present invention. Detailed Implementation

[0024] 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 some embodiments of the present invention, but not all embodiments.

[0025] This invention discloses a method for grid support and scheduling of mobile energy storage vehicle clusters, referring to... Figure 1 This includes steps S1-S4: S1. Obtain real-time parameters to form an adjustable output force upper limit vector.

[0026] In an optional embodiment, the scheduling platform establishes a remote communication connection with the battery management systems of each vehicle and continuously receives telemetry data frames reported by each mobile energy storage vehicle via the MQTT protocol according to a preset collection cycle. With a collection cycle of 1 second, the telemetry data frames are parsed to extract the real-time parameters of each vehicle's battery management system. Specifically, two key parameters are extracted from the telemetry messages at the battery stack level of each vehicle: the current state of charge (SOC) of the battery stack and the average internal resistance of the battery stack. The SOC is used to query the open-circuit voltage characteristic curve of the vehicle model, and the average internal resistance reflects the current discharge impedance level of the battery stack. Specifically, the SOC is taken from the `runtime.soc.mass` field of the telemetry message, and the average internal resistance is taken from the `runtime.r.mass` field. The collection results are written to the platform's time-series database at 1-second intervals, providing a continuously rolling real-time benchmark for subsequent power output limit calculations.

[0027] Furthermore, the internal resistance value stored in the runtime.r.mass field The equivalent internal resistance value output by the BMS after temperature-internal resistance characteristic compensation at the current temperature has been built into the BMS hardware layer for temperature correction. It can be used directly on the platform side without secondary compensation.

[0028] Next, for the first in the cluster For each vehicle, based on the current state of charge of the battery stack, the system queries the factory-calibrated open-circuit voltage characteristic curve of that vehicle model, which has been pre-imported into the platform database, to obtain the corresponding open-circuit voltage of the battery stack. The terminal voltage and discharge current satisfy the following battery equivalent circuit relationship:

[0029] In the formula, Indicates the first Real-time terminal voltage of the battery stack of the mobile energy storage vehicle; Indicates based on the first The open-circuit voltage of the battery stack obtained from the current state of charge query of the mobile energy storage vehicle; Indicates the first The current discharge current of the mobile energy storage vehicle; Indicates the first The average internal resistance of the battery stack in a mobile energy storage vehicle.

[0030] make This is equal to the DC bus undervoltage protection setting of the vehicle's PCS. Calculate the difference between the open-circuit voltage of the battery stack and the DC bus undervoltage protection setting value, and divide the difference by the average internal resistance of the battery stack to obtain the critical current that just reaches the DC bus undervoltage protection setting value:

[0031] In the formula, Indicates the first The critical current corresponding to the terminal voltage of the mobile energy storage vehicle when it drops to the DC bus undervoltage protection setting value; Indicates based on the first The open-circuit voltage of the battery stack obtained from the current state of charge query of the mobile energy storage vehicle; This indicates the preset DC bus undervoltage protection setting for the power AC / DC conversion equipment; Indicates the first The average internal resistance of the battery stack in a mobile energy storage vehicle.

[0032] The DC bus undervoltage protection setting for this PCS model is determined by the equipment's factory configuration file and is an inherent parameter of the equipment. When the DC bus voltage drops below... When the PCS triggers derating protection, the corresponding cloud platform alarm identifier is AI-PCS-DC-V004. Multiplying the critical current by the DC bus undervoltage protection setting and then by the preset inverter efficiency of the AC / DC power conversion equipment yields the adjustable output limit for each vehicle converted to the AC side without triggering derating protection.

[0033] In the formula, Indicates the first The adjustable output limit of the mobile energy storage vehicle converted to the AC side without triggering the derating protection. This indicates the preset DC bus undervoltage protection setting for the power AC / DC conversion equipment; Indicates the first The critical current corresponding to the terminal voltage of the mobile energy storage vehicle when it drops to the DC bus undervoltage protection setting value; This represents the inverter efficiency of the power AC / DC conversion device (e.g., 0.95). Indicates based on the first The open-circuit voltage of the battery stack obtained from the current state of charge query of the mobile energy storage vehicle; Indicates the first The average internal resistance of the battery stack in a mobile energy storage vehicle.

[0034] For all within the cluster Calculate each vehicle individually The final summation forms the upper limit vector of adjustable output force at the current moment. This vector serves as the core input for power allocation in this response event. It is updated dynamically each time a new BMS telemetry frame is received, ensuring that it reflects the current real-time battery physical state, rather than historical estimates.

[0035] Thus, by combining the temperature-compensated internal resistance data reported in real time by the BMS with the OCV-SOC calibration curve, the critical undervoltage derivation of the battery equivalent circuit model is directly embedded into the power output limit calculation stage. This allows the safe discharge boundary of each vehicle to be accurately quantified before the event is triggered, providing real-time constraint input with internal resistance as the core dimension for subsequent power allocation. This improves the inherent defect of overestimating the power output limit of high internal resistance vehicles when relying solely on the state of charge of the battery stack to estimate the discharge capacity due to ignoring the difference in internal resistance.

[0036] S2, Assign initial power command based on the upper limit vector of output.

[0037] In an optional embodiment, when the platform receives a response invitation issued by the grid side through the ancillary service market, it determines the total target power that the cluster needs to output or absorb based on the event type. The platform first sums the adjustable output limits of all vehicles in the adjustable output limit vector to obtain the maximum safe power of the cluster.

[0038] In the formula, This indicates the maximum safe power of the mobile energy storage vehicle cluster under its current healthy state; This indicates the total number of mobile energy storage vehicles participating in the current near real-time response event; Indicates the first The adjustable output limit of the mobile energy storage vehicle converted to the AC side without triggering the derating protection.

[0039] and the maximum safe power With total target power A size comparison was performed to verify the physical accessibility of the declared power. The comparison result reflects the maximum safe power under the current healthy state of the fleet. If... If the platform generates an early warning notification indicating insufficient response capability, it will send it to the power grid dispatching side and set the maximum safe power. As an internal execution target for this incident, the maximum safe power will be noted in the settlement declaration materials after the incident ends, and evidence for compliance defense will be retained.

[0040] After determining the total target power or internal execution target for the current response, calculate the ratio of the adjustable output limit of each vehicle to the maximum safe power. Combined with the margin factor, the initial power command allocated to each vehicle is determined according to the following proportional relationship:

[0041] In the formula, Indicates allocation to the first The current power command executed by the mobile energy storage vehicle; This indicates the total target power of the response events issued by the grid side; Indicates the first The adjustable output limit of the mobile energy storage vehicle converted to the AC side without triggering the derating protection. This indicates the maximum safe power of the mobile energy storage vehicle cluster under its current healthy state; This represents the margin coefficient determined based on historical reduction rates. This is used to actively reserve output margin during power distribution.

[0042] The ratio, the total target power, and the margin coefficient are multiplied together to obtain the initial power command for each vehicle, which is then issued for execution. This margin coefficient... When allocating full load, a certain percentage of output margin is proactively reserved for each vehicle. Since a fleet typically has a certain percentage of vehicles with reduced capacity during each response event, this embodiment... The value is determined based on the platform's historical response data: statistical data from the platform database shows the cluster's recent... The ratio of the number of vehicles that experienced PCS deductions in a given response event to the total number of vehicles participating in that event. The average of these events yields the historical rate of depreciation. To ensure that the initial general instructions meet the settlement red line, a lower limit value for the margin coefficient is set, letting... .in The value should cover at least one complete settlement cycle, for example, The value can be 12.

[0043] Next, the initial power command is sent to each vehicle's PCS via the platform's remote control channel. The PCS's remote_control field is set to true, the operating mode is switched to AC constant power mode (PQ mode), and the work_mode field is set to 6. The platform records the declared power, the power allocated to each vehicle, and the event start timestamp for this event, serving as baseline data for subsequent valid response verification. Communication methods capable of remote power command issuance are not limited to MQTT; those skilled in the art can select different downlink control protocols based on actual network conditions, and this invention does not limit this choice.

[0044] In this way, the adjustable output limit derived from the current internal resistance of each vehicle is used as the power allocation weight. Vehicles with a larger output limit undertake more tasks, while vehicles with a smaller output limit undertake fewer tasks. The actual commands of each vehicle strictly do not exceed the current safety boundary, reducing the risk of a single vehicle triggering PCS derating from the source, while also ensuring a margin coefficient. The introduction of this provides a basis for reserving non-zero margin for subsequent power transfer.

[0045] S3. Identify vehicles with reduced power limits and redistribute the power shortfall.

[0046] In an optional embodiment, during the response event execution, the platform continuously receives the real-time alarm status of each vehicle's power AC / DC conversion device (PCS), receiving the real-time status of the standard_id field in the PCS alarm array at a frequency of 1 second. It checks whether a DC bus undervoltage alarm flag appears in the real-time alarm status; if the alarm field of AI-PCS-DC-V004 corresponding to alarm_id changes from false to true, it simultaneously checks the AC side power derating alarm flag. Taking AI-PCS-AC-P001 as an example, it checks whether this flag is set; if the AC side power derating alarm flag is set simultaneously, the corresponding vehicle is ultimately identified as a derating vehicle, thus doubly confirming that the vehicle has indeed entered a derating state and eliminating single false alarms caused by voltage transient disturbances.

[0047] Furthermore, after confirming the reduction, the vehicle with the reduced amount is read. Measured power This measured value is read in real time from the ac_runtime.power.active.total field of the PCS AC side, ensuring that the power deficit calculation is based on actual collected values ​​rather than model estimations. The current executed power command for the derated vehicle is then calculated. The difference between the measured power and the actual power is used to obtain the power gap:

[0048] In the formula, The vehicle that has undergone derating protection. The corresponding power gap; Indicates the reduced vehicle The current power instruction being executed; Indicates the reduced vehicle The measured AC output power after derating.

[0049] It should be noted that the current power command in this step is the initial power command allocated in step S2 during the first power transfer; when a power transfer has already occurred, it is the power command that the corresponding vehicle updated and is currently executing after the previous power transfer.

[0050] Next, the platform will process the remaining normally operating vehicles that have not experienced a reduction in their credit limits. ( ), calculate the difference between the adjustable output limit of a normally operating vehicle and the corresponding currently executed power command, and obtain the available margin of the normally operating vehicle:

[0051] In the formula, This indicates the remaining normally operating vehicles that have not experienced a reduction in their depreciation rate. ( Available margin; Indicates the normally operating vehicle The adjustable output limit; Indicates the normally operating vehicle The current power instruction being executed.

[0052] Calculate the proportion of the available margin of each normally operating vehicle in the total available margin of all normally operating vehicles, weight the power gap based on the proportion, and add it to the corresponding currently executed power command:

[0053] In the formula, This indicates that after weighted allocation, the data is regenerated and distributed to the normally operating vehicles. The updated power command; Indicates the normally operating vehicle The current power instruction being executed; The vehicle that has undergone derating protection. The corresponding power gap; This indicates the remaining normally operating vehicles that have not experienced a reduction in their depreciation rate. ( Available margin; This represents the total available capacity of all normally operating vehicles.

[0054] It should be noted that the calculations obtained under extreme operating conditions... Greater than At that time, directly limit its amplitude to Issue and implement.

[0055] Vehicles with larger margins can handle more transfers, while vehicles with smaller margins can handle less. After the transfer, the command values ​​of each vehicle are still affected. The constraints will not trigger a new cascading reduction. After completing the above power transfer calculations, the updated power command is generated and redistributed to the PCS remote control interface of each normally operating vehicle. The entire data flow of detection-calculation-redistribution is guaranteed by MQTT QoS1 level latency, ensuring that power compensation is completed within a 15-minute response window.

[0056] In this way, by using the dual confirmation triggering method of DC bus undervoltage alarm and AC side derating alarm, combined with weighted power transfer calculation based on reserved margin, the actual total output of the cluster remains above 50% of the declared power after the derating event occurs, meeting the valid recognition conditions for near real-time response and effectively avoiding the risk that the derating of a single vehicle will drag down the response effectiveness of the entire cluster.

[0057] S4. Online internal resistance identification updates the upper limit and removes low-availability vehicles.

[0058] In an optional embodiment, after the response event ends, the platform extracts the measured electrical data of each vehicle for online internal resistance identification to update the adjustable output limit. Specifically, the platform reads the terminal voltage sampling sequence and discharge current sampling sequence of each vehicle during the execution of the response event, and extracts the time period in the discharge current sampling sequence where the relative change amplitude of the discharge current is lower than the steady-state determination threshold as the steady-state discharge segment. The relative change amplitude of the discharge current is the ratio of the absolute value of the current change between adjacent sampling points in the discharge current sampling sequence to the average discharge current of the discharge current sampling sequence, in order to eliminate the interference of startup transients and derating periods on the internal resistance estimation. The steady-state determination threshold is obtained as follows: Historical discharge current sequences from historical events of the same vehicle model are acquired, and the absolute values ​​of current changes between adjacent sampling points are extracted. A preset percentile of the absolute value sequence is calculated (e.g., the 75th percentile, selected to filter out the top 25% of abnormal current spikes caused by grid disturbances, retaining the majority of stable fluctuation characteristics). Simultaneously, the average discharge current of the historical discharge current sequence is calculated. The preset percentile and the average discharge current are divided to obtain the steady-state determination threshold for extracting the steady-state discharge segment. This threshold is archived with the vehicle model and does not require manual input.

[0059] Furthermore, during the steady-state discharge phase, since the current remains relatively stable, to eliminate transient polarization interference, the steady-state average terminal voltage and steady-state average discharge current during this period are directly calculated. Simultaneously, the state of charge (SOC) of the battery stack corresponding to this steady-state discharge phase is extracted (e.g., the average SOC or initial SOC during this period), and the open-circuit voltage characteristic curve is consulted to obtain the open-circuit voltage at this time. Subsequently, the updated average internal resistance of the battery stack is calculated using Ohm's law formula. This method directly utilizes measured electrical data generated by the event itself, without requiring dedicated internal resistance measurement excitation, and belongs to online operating condition identification. The data is written into the platform's energy storage device status database, replacing the original internal resistance estimate. Before the next response event is initiated, the platform recalculates based on the updated average internal resistance of the battery stack, ultimately obtaining the updated adjustable power limit, ensuring that the power allocation for each event is based on the latest physical state of the battery.

[0060] Next, based on the updated adjustable output upper limit distribution characteristics, vehicles in a low-availability state are identified and removed. This is for vehicles whose internal resistance has accumulated and increased after multiple events, leading to... For vehicles with a continuously decreasing output, the platform determines the following criteria: obtain the updated adjustable output limit of all registered vehicles in the cluster, and calculate the median. Extract the minimum adjustable output limit of normally operating vehicles within the cluster that did not experience derating protection in the historical response events, and calculate the ratio of the minimum value to the median value at the time the historical response event occurred; calculate the rolling average of the ratios from multiple historical response events to obtain the lower bound ratio of normally operating vehicle output. The median is multiplied by the ratio of the lower bound of the power output of normally operating vehicles to obtain the removal threshold. Finally, vehicles whose updated adjustable power output limit is lower than the removal threshold (i.e., current vehicles) are excluded from the removal process. If a vehicle is identified as low-availability at a certain time, it will be removed from the resource application list in the next scheduling cycle. The larger the value, the higher the overall health of the cluster, and the higher the removal threshold accordingly. The smaller the value, the lower the overall output level of the cluster, and the narrower the elimination threshold. This criterion is dynamically and adaptively adjusted according to the health of the fleet, without the need for manual intervention to set a fixed threshold.

[0061] In this way, the online internal resistance identification driven by the measured steady-state electrical data after the event, combined with the low availability status judgment criteria derived from the cluster output distribution characteristics, ensures that the output upper limit vector is refreshed in a timely manner after each event, continuously tracks the dynamic changes in the fleet health, and removes vehicles that are continuously judged as low availability from the resource application list, ensuring the overall response reliability of the cluster and avoiding the dragging down the monthly standby capacity settlement of the entire cluster by a single low availability vehicle.

[0062] Figure 2 This is a line graph comparing the actual output deviation of the cluster for each month of the year in this embodiment of the invention. The solid line marked with a solid circle represents the trend of actual output deviation of the existing conventional scheduling method; the dashed line marked with a hollow square represents the trend of actual output deviation of the method of the present invention. The absolute height of the data points on the curve represents the degree to which the actual output power of the cluster deviates from the target power, and the overall fluctuation of the curve represents the tendency of the deviation to deteriorate over time.

[0063] The image shows that the dashed line marked with hollow squares is not in the high deviation range of over 8%, but rather remains closely and steadily near the low benchmark of 3% to 4%. This proves that the technical solution of this invention successfully introduces the critical undervoltage derivation into the power distribution stage, eliminating the power loss interference caused by frequent undervoltage derating triggered by high internal resistance vehicles under high current loads, and without erroneously overestimating the actual discharge capacity of vehicles with deteriorated health. Observing the curve trend along the horizontal axis, it can be found that as the number of statistical months increases, the solid line marked with solid circles shows a clear oscillating upward trend; while the dashed line marked with hollow squares remains straight and does not diverge in output deviation with the extension of the overall fleet service time. This corresponds to the dynamic closed-loop control feature described in the specific implementation method, which continuously tracks the fleet health based on online internal resistance identification and adaptively eliminates vehicles in low availability.

[0064] It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this invention, and these modifications and improvements are all within the scope of protection of this invention.

Claims

1. A method for grid-supported dispatching of mobile energy storage vehicle clusters, characterized in that, include: S1: Obtain the real-time parameters of the battery management system of each vehicle in the mobile energy storage vehicle cluster, combine them with the preset DC bus undervoltage protection setting value, calculate the adjustable output limit of each vehicle under the condition of not triggering derating protection, and summarize them to form an adjustable output limit vector. S2: Calculate the maximum safe power of the cluster based on the adjustable output upper limit vector, combine the total target power issued by the grid side and the margin coefficient determined based on the historical derating rate, allocate initial power commands to each vehicle and issue them for execution; S3: During the response event execution, monitor the alarm status of each vehicle to identify the derated vehicle, calculate the power gap corresponding to the derated vehicle, and distribute the power gap based on the available margin of normally operating vehicles, generate an updated power command, and reissue it; S4 After the response event ends, the measured electrical data of each vehicle is extracted for online internal resistance identification to update the adjustable output limit. Based on the updated adjustable output limit distribution characteristics, vehicles in low availability status are determined and removed from the resource application list in the next scheduling cycle.

2. The method for grid support and dispatching of mobile energy storage vehicle clusters according to claim 1, characterized in that, The process of acquiring real-time parameters of the battery management system of each vehicle in the mobile energy storage vehicle cluster, combined with a preset DC bus undervoltage protection setting, calculates the adjustable output limit of each vehicle under the condition of not triggering derating protection, and summarizes them to form an adjustable output limit vector. This includes: extracting the current state of charge of the battery stack and the average internal resistance of the battery stack for each vehicle; querying the open-circuit voltage characteristic curve based on the current state of charge of the battery stack to obtain the corresponding open-circuit voltage of the battery stack; calculating the difference between the open-circuit voltage of the battery stack and the DC bus undervoltage protection setting, and dividing the difference by the average internal resistance of the battery stack to obtain the critical current; multiplying the critical current by the DC bus undervoltage protection setting and multiplying by the preset inverter efficiency of the AC-DC power conversion equipment to obtain the adjustable output limit of each vehicle converted to the AC side under the condition of not triggering derating protection, and summarizing them to form an adjustable output limit vector.

3. The method for grid support and dispatching of mobile energy storage vehicle clusters according to claim 1, characterized in that, The process of calculating the maximum safe power of the cluster based on the adjustable output upper limit vector, combining the total target power issued by the grid side and the margin coefficient determined based on the historical derating rate, and allocating and executing initial power commands to each vehicle includes: summing the adjustable output upper limits of all vehicles in the adjustable output upper limit vector to obtain the maximum safe power of the cluster; statistically analyzing the proportion of vehicles that experienced derating in historical response events as the historical derating rate, and determining the margin coefficient based on the historical derating rate and a preset lower limit value of the margin coefficient; calculating the ratio of the adjustable output upper limit of each vehicle to the maximum safe power; if the maximum safe power is not less than the total target power, multiplying the ratio, the total target power, and the margin coefficient together; if the maximum safe power is less than the total target power, replacing the total target power with the maximum safe power and multiplying together, finally obtaining the initial power command corresponding to each vehicle and executing it.

4. The method for grid support and dispatching of mobile energy storage vehicle clusters according to claim 1, characterized in that, The monitoring of the alarm status of each vehicle to identify derating vehicles includes: continuously receiving the real-time alarm status of the AC / DC power conversion equipment contained in each vehicle; detecting whether a DC bus undervoltage alarm flag appears in the real-time alarm status; if the DC bus undervoltage alarm flag appears, simultaneously checking whether the AC side power derating alarm flag is set; if the AC side power derating alarm flag is set simultaneously, the corresponding vehicle is finally identified as the derating vehicle.

5. A method for grid support and dispatching of mobile energy storage vehicle clusters according to claim 1, characterized in that, The process of calculating the power gap corresponding to the derated vehicle, allocating the power gap based on the available margin of normally operating vehicles, generating an updated power command, and reissuing it includes: reading the measured power of the derated vehicle, calculating the difference between the current executed power command of the derated vehicle and the measured power to obtain the power gap; calculating the difference between the adjustable output limit of the normally operating vehicle and the corresponding current executed power command to obtain the available margin of the normally operating vehicle; calculating the proportion of the available margin of each normally operating vehicle in the total available margin of all normally operating vehicles; weighting the power gap based on the proportion and adding it to the corresponding current executed power command; if the sum is greater than the corresponding adjustable output limit, limiting it to the adjustable output limit, and finally generating the updated power command and reissuing it.

6. A method for grid support and dispatching of mobile energy storage vehicle clusters according to claim 2, characterized in that, The step of extracting measured electrical data from each vehicle for online internal resistance identification to update the adjustable output limit includes: reading the terminal voltage sampling sequence and discharge current sampling sequence of each vehicle during the execution of the response event; extracting the time period from the discharge current sampling sequence where the relative change amplitude of the discharge current is lower than the steady-state determination threshold as the steady-state discharge segment; within the steady-state discharge segment, obtaining the battery stack state of charge corresponding to the steady-state discharge segment to query the open-circuit voltage at this time; and calculating the updated battery stack internal resistance average value based on the open-circuit voltage and the steady-state average value of the terminal voltage sampling sequence and the discharge current sampling sequence according to Ohm's law formula; and recalculating based on the updated battery stack internal resistance average value to finally obtain the updated adjustable output limit.

7. A method for grid support and dispatching of mobile energy storage vehicle clusters according to claim 6, characterized in that, The step of extracting the time period in which the relative change of discharge current is lower than the steady-state determination threshold from the discharge current sampling sequence as the steady-state discharge segment includes: obtaining the historical discharge current sequence from historical events of the same vehicle model; calculating the preset percentile of the absolute value of the current change between adjacent sampling points of the historical discharge current sequence; calculating the average discharge current of the historical discharge current sequence; and performing a division operation between the preset percentile and the average discharge current to finally obtain the steady-state determination threshold to extract the steady-state discharge segment.

8. A method for grid support and dispatching of mobile energy storage vehicle clusters according to claim 1, characterized in that, The process of determining low-availability vehicles based on the updated adjustable output limit distribution characteristics and removing them from the resource application list in the next scheduling cycle includes: obtaining the updated adjustable output limit of all registered vehicles in the cluster and calculating the median; extracting the minimum adjustable output limit of normally operating vehicles in the cluster that have not experienced derating protection in historical response events and calculating the ratio of the minimum value to the median at the time of the historical response event; calculating the rolling average of the ratios from multiple historical response events to obtain the lower bound ratio of normally operating vehicle output; multiplying the median by the lower bound ratio of normally operating vehicle output to obtain the removal threshold; and finally determining vehicles whose updated adjustable output limit is lower than the removal threshold as low-availability vehicles and removing them.

9. A method for grid support and dispatching of mobile energy storage vehicle clusters according to claim 1, characterized in that, After calculating the maximum safe power of the cluster based on the adjustable output upper limit vector, the method further includes: comparing the maximum safe power with the total target power; if the maximum safe power is less than the total target power, generating and sending an early warning notification of insufficient response capability; using the maximum safe power as the internal execution target of this event, and attaching the maximum safe power to the settlement declaration materials after the event ends.

10. A method for grid support and dispatching of mobile energy storage vehicle clusters according to claim 1, characterized in that, The process of acquiring real-time parameters of the battery management systems of each vehicle in the mobile energy storage vehicle cluster includes: establishing a remote communication connection with the battery management systems of each vehicle; continuously receiving telemetry data frames reported by each vehicle according to a preset acquisition cycle; parsing the telemetry data frames; and finally extracting the real-time parameters of the battery management systems of each vehicle.

Images