An external EMS energy management system

Abstract

The application is suitable for the technical field of energy management system, and provides an external EMS energy management system, which is independent of the energy storage converter, and is in communication connection with the gateway power meter and the energy storage converter, and comprises a gateway power acquisition module, an inversion state acquisition module, a time sequence analysis module, a control decision module, a statistical analysis module and a fault diagnosis module. The gateway power acquisition module is used for acquiring time sequence data of the gateway active power according to a preset sampling period. The inversion state acquisition module is used for acquiring the running state of the energy storage converter, and the single-point threshold judgment of the existing external EMS is upgraded to a joint judgment based on the time sequence trend, the inversion state and the battery state, and a complete closed loop is formed through battery cooperation, adaptive sampling and multiple event diagnosis, which improves the accuracy of the anti-inversion control, reduces the misjudgment and omission, improves the utilization rate of the energy storage, enhances the stability of the system and reduces the operation and maintenance difficulty.

Description

Technical Field

[0001] This invention relates to the field of external EMS energy management system technology, and more specifically, to an external EMS energy management system. Background Technology

[0002] With the widespread application of distributed photovoltaic and energy storage systems, there is a common risk of power backflow to the grid at the user-side grid connection point. To prevent backflow, external EMS (Energy Management System) is usually set up independently of the energy storage converter. It collects the active power at the point of sale by the meter and adjusts the charging and discharging power of the energy storage converter accordingly to keep the power at the grid connection point within the allowable range.

[0003] The existing anti-backflow control method of external EMS is generally as follows: read the current value of active power at the gate meter at a fixed sampling period, compare the current value with the preset anti-backflow threshold; when the current value of active power at the gate is lower than the anti-backflow threshold, limit the discharge power of the energy storage converter to prevent power from being fed back to the grid.

[0004] However, existing anti-backflow measures rely solely on the current sampled value of the threshold power for single-point judgment, failing to identify the changing trend of the threshold power. Instantaneous load fluctuations, short-term changes in photovoltaic output, or sampling noise can all cause the threshold power to briefly fall below the threshold at a certain sampling point, leading to malfunctions. Furthermore, these anti-backflow measures can only perform single-event anti-backflow control and lack the ability to analyze multiple backflow events. When the system repeatedly experiences backflow over a period of time, maintenance personnel find it difficult to quickly determine the source of the problem, resulting in low troubleshooting efficiency. To address these issues, an external EMS energy management system is proposed. Summary of the Invention

[0005] In view of the shortcomings of the existing technology, the purpose of this invention is to provide an external EMS energy management system.

[0006] To achieve the above objectives, the present invention provides the following technical solution: An external EMS (Energy Management System) is provided, which is independent of the energy storage converter and communicatively connected to the gateway meter and the energy storage converter. The external EMS includes a gateway power acquisition module, an inverter status acquisition module, a timing analysis module, a control decision module, a recovery judgment module, a battery status acquisition module, a battery management coordination module, an adaptive sampling control module, a reverse current event recording module, a statistical analysis module, and a fault diagnosis module. All modules are located within the external EMS and interact with each other and transmit commands via an internal data bus or software interface.

[0007] The output of the gate power acquisition module is connected to the input of the timing analysis module. The output of the timing analysis module is connected to the input of the control decision module. The output of the inverter state acquisition module is connected to the input of the control decision module. The output of the battery state acquisition module is connected to the inputs of both the control decision module and the battery management coordination module. The output of the control decision module is connected to the input of the battery management coordination module. The output of the battery management coordination module is connected to the battery management system.

[0008] The threshold power acquisition module is used to acquire time-series data of the threshold active power according to a preset sampling period. The inverter status acquisition module is used to acquire the operating status of the energy storage converter, at least distinguishing between the inverter state and the non-inverter state. The time-series analysis module is used to perform trend analysis on the threshold active power time-series data within a sliding time window and extract the threshold power change characteristics. The control decision module is used to determine whether there is a reverse current risk when the energy storage converter is in the inverter state, based on the threshold power change characteristics, and generate an anti-reverse current control command when a reverse current risk is determined to exist. The battery management coordination module is used to send a battery-side power limiting command. The recovery judgment module is used to release the anti-reverse current limit based on the threshold power change characteristics.

[0009] The present invention is further configured such that the gate power change feature includes at least: current gate power, window average change rate, predicted gate power, and the number of sampling points within the window that are consecutively below a first threshold. The window average change rate is the difference between the gate power at the end of the window and the gate power at the beginning of the window, divided by the window time span. The predicted gate power is the current gate power plus the product of the window average change rate and a preset lead time.

[0010] The present invention is further configured such that the control decision module determines the existence of backflow risk when the following conditions are met simultaneously: The energy storage converter is in inverter mode; the average rate of change of the window is less than the first rate of change threshold; the current threshold power is less than or equal to the first threshold, or the predicted threshold power is less than or equal to the first threshold; the number of sampling points within the window that are continuously below the first threshold reaches the preset number of times condition.

[0011] The present invention is further configured such that the recovery judgment module releases the anti-backflow restriction when the following conditions are simultaneously met: the current gate power is greater than a second threshold, and this restriction persists for a preset number of sampling periods; the average rate of change of the window is greater than a second rate of change threshold. Wherein, the second threshold is greater than the first threshold.

[0012] The present invention is further configured such that: the battery status acquisition module is used to obtain the maximum allowable discharge power of the battery from the battery management system; when judging the reverse current risk, the control decision module also makes a joint judgment based on the maximum allowable discharge power of the battery; when the current actual discharge power of the energy storage converter exceeds the maximum allowable discharge power of the battery, it does not determine that there is a reverse current risk or reduces the reverse current risk level; while generating the anti-reverse current control command, the battery management coordination module sends a battery-side power limiting command to the battery management system, so that the energy storage converter and the battery management system act in coordination.

[0013] The present invention is further configured such that the battery-side power limiting command includes at least one of the following: The upper limit of the battery's allowable discharge power is temporarily adjusted to the target value of the power limit.

[0014] When the power limit target value is zero or negative, the battery management system is instructed to switch the battery side to charging or standby mode.

[0015] Adjust the upper limit of the battery discharge current so that the battery output power does not exceed the target power limit value.

[0016] The present invention is further configured such that the power limiting target value is the current gate power minus the first threshold.

[0017] When the power limit target value is positive, the energy storage converter is controlled to discharge at a power not exceeding the power limit target value.

[0018] When the power limit target value is zero or negative, the energy storage converter is controlled to switch to standby or charging state.

[0019] The invention is further configured such that the output of the adaptive sampling control module is connected to both the gate power acquisition module and the time series analysis module. The adaptive sampling control module dynamically adjusts the sampling period and the sliding window length based on a risk margin index. The risk margin index is calculated based on the difference between the current gate power and a first threshold, as well as the average rate of change of the window. A larger risk margin index results in a longer sampling period and a smaller window length. Conversely, a smaller risk margin index results in a shorter sampling period and a larger window length.

[0020] The invention is further configured such that: the fault diagnosis module is used to determine the location of the backflow problem through a hierarchical decision tree based on event characteristic parameters and statistical characteristics; the backflow event recording module is used to record, within a statistical time period, situations where the backflow risk judgment condition is met or the actual gate power is less than a first threshold for at least one sampling cycle as a backflow event, and to record the event characteristic parameters; the statistical analysis module is used to aggregate and statistically analyze multiple backflow events within the statistical time period to obtain statistical characteristics.

[0021] The present invention is further configured such that the hierarchical decision tree includes: The first layer: Determine whether there is a reverse current event in the non-inverter state of the energy storage converter. If so, determine whether it is photovoltaic independent backfeed, load feedback or metering sampling abnormality based on the relationship between photovoltaic power and load power.

[0022] The second layer: Determine whether the average response delay of the energy storage converter is greater than the preset response delay threshold. If so, it is determined to be a problem on the communication control side.

[0023] The third layer: Determine whether the power change rate of the energy storage converter is close to the maximum allowable value and whether the power at the cut-off point is still lower than the first threshold during the event. If so, proceed to the fourth layer.

[0024] The fourth layer: Determine the correlation between the event and the sudden drop in load power or the sudden increase in photovoltaic power, and determine whether the change is too rapid on the load side or the photovoltaic side.

[0025] Fifth layer: Determine whether the event duration is extremely short and whether there is an abnormal jump in the gate power. If so, it is determined to be an abnormal metering sampling; otherwise, it is determined to be an unreasonable control parameter setting.

[0026] In summary, this application includes at least one of the following beneficial technical effects: (1) By performing trend analysis on the gate power within the sliding time window through the time series analysis module, the average change rate of the window, the predicted gate power and the number of consecutive times below the threshold are extracted. This can identify the backflow trend in advance, issue an early warning and execute anti-backflow control before the gate power crosses the threshold, significantly reducing the missed judgment caused by judgment lag. At the same time, by requiring multiple consecutive sampling points to be below the threshold before triggering the action, the misjudgment caused by instantaneous load fluctuations and sampling noise is effectively filtered out, improving the accuracy of anti-backflow control.

[0027] (2) By introducing an inverter status acquisition module, the inverter status of the energy storage converter is taken as a necessary condition for determining the reverse flow risk, which reduces the false operation caused by the power drop at the switch in the non-inverter state and improves the effective utilization rate of the energy storage system. Furthermore, the reverse flow event recording module and statistical analysis module perform structured recording and aggregated statistics on multiple reverse flow events, and the fault diagnosis module uses a hierarchical decision tree to automatically determine the problem location, which can quickly distinguish the causes such as load-side sudden change, photovoltaic-side sudden change, communication control lag, metering sampling abnormality or unreasonable control parameters, reduce the difficulty of operation and maintenance investigation, and improve the system operation and maintenance efficiency.

[0028] (3) By using the battery status acquisition module and the battery management coordination module, the battery-side power limiting command is sent to the battery management system at the same time during the anti-reverse current control, and the discharge power is limited from the DC side, making the power limiting process of the energy storage converter more stable, reducing the risk of sudden changes in DC bus voltage and malfunction of the battery management system protection, and improving the system operation stability.

[0029] (4) The risk margin index is calculated by the adaptive sampling control module. The sampling period and window length are dynamically adjusted according to the risk level. The sampling frequency and window density are increased when the risk is high to ensure the response speed, and the sampling frequency is reduced when the risk is low to save communication bandwidth and computing resources, thus realizing a dynamic balance between system performance and operating cost. Attached Figure Description

[0030] Figure 1 This is a schematic diagram of the overall structure of an external EMS energy management system according to the present invention.

[0031] Figure 2 This is a schematic diagram of the functional module structure built into the external EMS controller in this invention.

[0032] Figure 3 This is a schematic diagram illustrating the three states constituted by single-current risk assessment and recovery control in this invention.

[0033] Figure 4 This is a schematic diagram of the hierarchical decision tree for diagnosing multiple reverse flow faults in this invention.

[0034] Figure 5 This is a schematic diagram of the overall workflow in this invention.

[0035] Figure 6 This is a schematic diagram of the adaptive sampling control rules in this invention. Detailed Implementation

[0036] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0037] It should be noted that, unless otherwise specified, all technical and scientific terms used in this application have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0038] Please see Figures 1-6 The present invention provides the following technical solutions: Example 1, see Figure 1 An external EMS energy management system includes an external EMS controller, and a gateway meter, an energy storage converter, and a battery management system that are communicatively connected to the external EMS controller.

[0039] The external EMS controller is set up independently of the energy storage converter. It obtains the active power data of the gateway meter, the operating status of the energy storage converter, and the battery status parameters of the battery management system through the communication bus, and sends power control commands to the energy storage converter and the battery management system.

[0040] In practical engineering, energy storage converters come from various brands and use different communication protocols. If an external EMS relies on data from the internal components of the energy storage converter, it will face the risks of interface compatibility difficulties and data asynchrony. Power at the switching point is the true indicator of whether reverse current has occurred. Direct acquisition by an external EMS can reduce communication delays and errors caused by internal processing within the energy storage converter. Meanwhile, inverter status is a necessary prerequisite for reverse current prevention, but a single judgment method may lead to misjudgments due to communication interruptions or signal anomalies; therefore, multiple redundant acquisition methods are needed.

[0041] When the external EMS is set up independently of the energy storage converter, the basic data required to determine the reverse current can be accurately obtained by setting the overall architecture of the external EMS and its specific functional modules, and its functional modules can be used to identify the inverter status of the energy storage converter.

[0042] See Figure 2 The external EMS controller includes the following functional modules: The threshold power acquisition module is used to acquire time-series data of the threshold active power according to a preset sampling period, forming a sliding time window sequence; the inverter status acquisition module is used to acquire the operating status of the energy storage converter, at least distinguishing between the inverter state and the non-inverter state, and outputting binary state features; the battery status acquisition module is used to acquire at least one of the following from the battery management system: the current state of charge of the battery, the maximum allowable discharge power of the battery, the maximum allowable charging power of the battery, and the battery temperature; the time-series analysis module is used to perform forward and backward trend analysis on the threshold active power sequence within the sliding time window and extract the threshold power change features; the control decision module is used to determine whether there is a reverse current risk based on the threshold power change features and the inverter status, using a reverse current risk discrimination model, and generate anti-reverse current measures when a risk exists. The system includes: a flow control command module; a battery management coordination module that generates a battery-side power limiting command based on battery state parameters when anti-reverse current control is activated, and adjusts the actual charging and discharging power of the battery through the battery management system to enable the energy storage converter and the battery management system to work together; a recovery judgment module that determines whether to lift the limit based on recovery conditions after anti-reverse current control is activated; an adaptive sampling control module that dynamically adjusts the sampling period and sliding window length based on risk margin indicators; a reverse current event recording module that records the characteristic parameters of each reverse current event within a statistical time period; a statistical analysis module that aggregates and statistically analyzes multiple reverse current events within a statistical time period; and a fault diagnosis module that determines the location of the reverse current problem through a hierarchical decision tree based on statistical and event characteristics and outputs the diagnostic results.

[0043] In the modules described above, the sampling period, sliding window length, and statistical time period are all configurable parameters. Typically, the sampling period can be selected from 100 milliseconds to 1 second, the sliding window length can be selected from 5 to 50, and the statistical time period is set to 10 minutes by default, which can be adjusted according to actual needs.

[0044] All of the above functional modules are located inside the external EMS controller, and data interaction and command transmission are carried out through the controller's internal data bus or software interface.

[0045] The specific connection relationships of the above modules are as follows: The input end of the gate power acquisition module is connected to the gate electricity meter, and the output end is connected to the time series analysis module, used to send the acquired gate active power time series data to the time series analysis module; the output end of the time series analysis module is connected to the control decision module and the adaptive sampling control module respectively, used to send the extracted gate power change characteristics to the control decision module, and send the window average change rate, current gate power, etc. to the adaptive sampling control module; the input end of the inverter status acquisition module is connected to the energy storage converter, and the output end is connected to the control decision module, used to send the inverter status binary characteristics to the control decision module; the input end of the battery status acquisition module is connected to the battery management system, and the output end is connected to the control decision module and the battery management coordination module respectively, used to send the battery's maximum allowable discharge power, state of charge, etc. to the control decision module and the battery management coordination module; the output end of the control decision module is connected to the energy storage converter, the battery management coordination module, and the reverse current event recording module respectively. The system is used to send anti-reverse current control commands to the energy storage converter, send trigger signals to the battery management coordination module, and notify the reverse current event recording module when a reverse current risk is detected. The output of the battery management coordination module is connected to the battery management system to send battery-side power limiting commands. The input of the recovery judgment module is connected to the timing analysis module, and the output is connected to the control decision module to notify the control decision module to lift the limit when the recovery conditions are met. The output of the adaptive sampling control module is connected to the gate power acquisition module and the timing analysis module respectively to dynamically adjust the sampling period and the sliding window length. The output of the reverse current event recording module is connected to the statistical analysis module to output the recorded event characteristic parameters to the statistical analysis module. The output of the statistical analysis module is connected to the fault diagnosis module to output the aggregated statistics to the fault diagnosis module. The output of the fault diagnosis module can be connected to the human-machine interface or remote monitoring platform and can be fed back to the control decision module or parameter configuration module to optimize control parameters.

[0046] The inverter status acquisition module determines whether the energy storage converter is in inverter state through at least one of the following methods: Read the operating mode flag bit in the status word of the energy storage converter. When the flag bit indicates "discharge" or "inverter", it is determined to be in inverter mode.

[0047] The three-phase voltage and current on the AC side of the energy storage converter are collected, and the active power on the AC side is calculated. When the active power is greater than the preset dead zone threshold, it is determined to be in inverter mode. When it is less than the negative preset dead zone threshold, it is determined to be in charging mode. Otherwise, it is determined to be in standby mode.

[0048] The DC side voltage and DC current of the energy storage converter are collected. When the DC current flows from the battery to the energy storage converter and the DC side voltage is higher than the AC side rectified voltage, it is determined to be in inverter mode.

[0049] Read the "operating direction" flag reported by the energy storage converter via communication.

[0050] When the results of multiple methods are inconsistent, the direction of active power on the AC side shall prevail, because the active power on the AC side directly reflects whether the energy storage converter outputs active power to the load or the grid, and can more accurately characterize the actual operating status.

[0051] The gate power timing data and inverter state binary features output in this embodiment will serve as inputs to the timing analysis module, conditions for the risk discrimination model, and the basis for the mid-trigger battery collaborative control, respectively. The basic data and state identification established through the above structure provide the foundation for the implementation of all subsequent functions.

[0052] Example 2: This example builds upon the time-series power data and inverter state binary features obtained in Example 1. It extracts features from the power sequence within the sliding window. In actual operation, load start-up and shutdown, and photovoltaic output fluctuations are frequent. The power threshold may briefly fall below the threshold at a certain sampling point due to instantaneous disturbances, leading to erroneous actions from single-point judgments. Furthermore, relying solely on the current value cannot predict future over-limit risks. Therefore, a sliding window mechanism is needed to extract features such as the rate of change, predicted value, and duration, upgrading single-point judgments to continuous linear judgments.

[0053] Specifically, the time series analysis module configures the settings. This module maintains a sliding time window, storing the active power data of the gate from the most recent sampling periods. Assuming the current sampling time is the latest sample, the time series analysis module extracts features from the power sequence within the window, obtaining the following features: Current gate power: the active power value of the gate obtained from the latest sampling; Window average rate of change: reflects the overall trend of change in gate power within the window. It is calculated by dividing the difference between the gate power at the end of the window and the gate power at the beginning of the window by the time span within the window. Predicted threshold power: Used to anticipate future risks before the current value crosses the threshold. It is calculated by adding the current threshold power to the product of the window's average rate of change and a preset lead time. The number of consecutive sampling points below the first threshold within the window: used to filter out instantaneous fluctuations; Inverter state binary characteristics: provided by the inverter state acquisition module, the state is taken as valid when the energy storage converter is in inverter state, and otherwise as invalid. Maximum allowable discharge power of the battery: This is obtained in real time from the battery management system by the battery status acquisition module and is used to jointly determine whether the battery has inverter capability.

[0054] The average rate of change of the window can also be the slope obtained by performing least-squares linear fitting on all sampling points within the window, in order to further reduce the interference of sampling noise on trend judgment.

[0055] The multidimensional feature vector extracted in this embodiment will be directly used as the input variable of the reverse flow risk discrimination model. Simultaneously, the adaptive sampling control module also needs to utilize the average window change rate and the current gate power output from the above structure to calculate the risk margin index.

[0056] Example 3, see Figure 3 In practical judgment, relying solely on the current power being below the threshold can lead to misjudgment, relying solely on the rate of change is susceptible to noise interference, relying solely on the inverter state fails to identify trends, and relying solely on the predicted value may result in accumulated errors. Therefore, based on the multi-dimensional features including current power, rate of change, predicted value, duration, and inverter state already obtained in Example 2, it is necessary to establish a parameterized joint discrimination model to integrate all conditions into a logical whole.

[0057] The control decision module employs a parameterized reverse flow risk discrimination model to comprehensively assess the aforementioned characteristics. The model involves the following preset parameters: a first threshold, a second threshold, a first rate of change threshold, a second rate of change threshold, a condition for the number of consecutive sampling points below the first threshold, and the number of continuous sampling periods for recovery.

[0058] The first threshold is the anti-backflow trigger threshold; the second threshold is the recovery threshold, and the second threshold is greater than the first threshold.

[0059] The first rate of change threshold is used to determine whether the power at the gate is showing a downward trend; The second rate of change threshold is used to determine whether the power at the threshold is showing an upward trend.

[0060] The logic for determining backflow risk is as follows: a backflow risk is determined to exist when the following conditions are met simultaneously: The energy storage converter is in inverter mode; The average rate of change of the window is less than the first rate of change threshold, which means that the power at the gate is decreasing. The current gate power is less than or equal to the first threshold, or the predicted gate power is less than or equal to the first threshold; The number of consecutive sampling points below the first threshold within the window reaches the preset number of times condition; The actual discharge power of the energy storage converter does not exceed the maximum allowable discharge power of the battery, meaning the battery has the ability to maintain the current inverter output.

[0061] A joint judgment condition for the maximum allowable discharge power of the battery is introduced. Specifically, when the battery discharge is limited due to low state of charge, abnormal temperature, or internal protection strategies of the battery management system, even if the threshold power trend meets the risk condition, the battery itself can no longer provide sufficient discharge power, and the actual reverse current risk may have been eliminated or reduced. In this case, limiting the energy storage converter would be redundant. By introducing this condition, the system can more accurately determine whether the energy storage converter has the capability to cause reverse current, thereby further reducing the probability of malfunction.

[0062] As an alternative, the battery state of charge can be introduced as a judgment condition: when the battery state of charge is lower than the preset lower limit, the battery is close to being discharged and the energy storage converter cannot maintain inversion. At this time, the inversion state can be regarded as invalid or the risk level can be reduced in advance to avoid invalid power limiting commands.

[0063] When a reverse current risk is detected, the control decision module generates an anti-reverse current control command. The anti-reverse current control command includes at least one of the following: reducing the discharge power of the energy storage converter; increasing the charging power of the energy storage converter; switching the energy storage converter from a discharge state to a charging state or a standby state; adjusting reactive power to reduce the risk of active power exceeding the limit at the threshold.

[0064] The power limit target value is determined as follows: subtract the first threshold from the current threshold power; when the difference is positive, control the energy storage converter to discharge at a power not exceeding the difference; when the difference is zero or negative, control the energy storage converter to switch to standby or charging mode to eliminate the reverse current risk as soon as possible.

[0065] The risk assessment result output in this embodiment directly determines the actions of subsequent modules: when a risk is determined to exist, battery-side collaborative control is triggered, which also serves as one of the triggering conditions for reversal event recording; when no risk is determined to exist, the recovery assessment process is initiated. Furthermore, the adaptive sampling control module also adjusts its sampling parameters based on the risk margin involved in this embodiment.

[0066] Example 4, see Figure 3When Example 3 determines that there is a reverse current risk and generates a power limiting command for the energy storage converter, if the battery side continues to discharge at the original power, the DC bus voltage will rise, potentially triggering the battery management system's overvoltage protection or even shutdown. If the external EMS only controls the energy storage converter without coordinating with the battery management system, the reverse current prevention process may cause new faults. Therefore, based on the risk assessment results of Example 3, this example introduces a battery management coordination module to synchronously issue a battery-side power limiting command to the battery management system, limiting the battery discharge power and making the power limiting process of the energy storage converter smoother, forming a closed loop of gate power detection, energy storage converter power limiting, and battery management system current limiting.

[0067] When a reverse current risk is detected, the control decision module generates a power limiting command for the energy storage converter, as well as a battery-side power limiting command, and sends it to the battery management system via the battery management coordination module. This battery-side command includes at least one of the following: The upper limit of the battery's allowable discharge power is temporarily adjusted to the aforementioned power limit target value, so that the energy storage converter cannot obtain power from the battery side exceeding the limit value; When the power limit target value is zero or negative, the battery management system is instructed to switch the battery side working state to charging or standby and prohibit discharge. Adjust the upper limit of the battery discharge current so that the battery output power does not exceed the above-mentioned power limit target value.

[0068] The coordinated operation sequence of the energy storage converter and the battery management system is as follows: the external EMS controller simultaneously sends limiting commands to both the energy storage converter and the battery management system; the energy storage converter begins to reduce the AC side output power; the battery management system simultaneously reduces the upper limit of the battery side discharge current, reducing the DC side power output; since the DC side power is reduced first, the AC side power of the energy storage converter decreases more smoothly, avoiding sudden changes in the DC bus voltage; when the recovery conditions are met, the external EMS simultaneously sends recovery commands to both the energy storage converter and the battery management system, gradually restoring normal charging and discharging.

[0069] Through the above-mentioned coordinated control, a complete closed loop is formed, which includes "power detection at the cutoff point, power limiting of the energy storage converter, current limiting of the battery management system, power recovery at the cutoff point, and gradual recovery".

[0070] In this embodiment, the status of the coordinated control action is fed back into the recovery judgment process: the recovery condition can only be met when both the energy storage converter and the battery management system execute power limiting commands and the gate power recovers. Simultaneously, in this embodiment, the actual response on the battery side (such as the battery's allowable discharge power and SOC changes) will be recorded as event characteristic parameters, providing a basis for battery-side judgment in fault diagnosis.

[0071] Example 5, see Figure 3After determining the risk in Implementation Example 3 and triggering the coordinated control in Implementation Example 4, when the reverse flow risk is eliminated and the threshold power recovers, it is necessary to determine the timing for lifting the restriction. If the recovery threshold is the same as or close to the trigger threshold, small fluctuations in the threshold power near the threshold will cause the energy storage converter to repeatedly switch between power limiting and recovery, resulting in system oscillation. Therefore, this embodiment sets a recovery threshold higher than the trigger threshold and requires the restriction to be lifted only when the threshold power is continuously higher than the recovery threshold and shows an upward trend, forming hysteresis control.

[0072] The recovery judgment module will lift the anti-backflow restriction when the following conditions are met simultaneously: The current gate power is greater than the second threshold and continues for a preset number of sampling periods; The average rate of change of the window is greater than the second rate of change threshold, which means that the power at the gate is showing an upward trend.

[0073] The second threshold is greater than the first threshold. By setting a hysteresis loop, frequent actions caused by fluctuations in gate power around the threshold can be effectively avoided.

[0074] The aforementioned single-time backflow risk assessment and recovery control together constitute three states: normal state, backflow prevention restriction state, and recovery waiting state.

[0075] The recovery conditions and state transition results in this embodiment will serve as one of the bases for determining the end of the reverse flow event: the current reverse flow event record can only end when the recovery conditions are met and the restrictions are lifted.

[0076] Meanwhile, this embodiment, together with Embodiments 3 and 4, constitutes a complete closed-loop control system for a single backflow event, enabling the system to act promptly when a risk occurs and to recover smoothly after the risk is eliminated. Furthermore, the parameters involved in the recovery judgment (such as the number of continuous sampling periods for recovery) are also one of the diagnostic criteria for whether the control parameter settings are reasonable.

[0077] Example 6: In Example 1, the sampling period and sliding window length are set to fixed values, but the backflow risk is not constant in actual operation. If the sampling parameters are set according to the worst-case scenario, the system will run at an excessively high frequency most of the time, wasting resources; if the sampling period is too long, it will be unable to capture trends in time when the risk changes rapidly. Therefore, this example uses the average window change rate and current threshold power extracted in Example 2, as well as the risk margin concept involved in Example 3, to calculate the risk margin index, and adjusts the sampling period and window length in stages according to the index size.

[0078] The adaptive sampling control module calculates a risk margin index based on the average rate of change of the window in the current features and the difference between the current gate power and the first threshold. The risk margin index is calculated by subtracting the first threshold from the current gate power and then adding the product of the average rate of change of the window and the preset lead time. The larger the index, the safer the current situation; the smaller the index, the closer to the risk of backflow.

[0079] See Figure 6 The adaptive rule is: When the risk margin index is greater than the preset high margin threshold, it is judged as low risk, the sampling period is adjusted to a slower value, the sliding window length is adjusted to a smaller value, and the calculation frequency is reduced. When the risk margin index is between the preset low margin threshold and the high margin threshold, it is judged as medium risk, and the sampling period and window length are kept at the middle value. When the risk margin index is less than or equal to the preset low margin threshold, it is judged as high risk. The sampling period is shortened to a faster value, the window length is increased, and the response speed is improved.

[0080] If the system remains in a low-risk state for more than a preset time, the sampling frequency can be further reduced to an even lower value to save resources to the maximum extent.

[0081] The dynamically adjusted sampling parameters in this embodiment will be fed back to the gate power acquisition module in Embodiment 1, affecting the sliding window data composition in Embodiment 2, and consequently affecting the feature calculation and risk discrimination sensitivity in Embodiment 3. Therefore, this embodiment is an adaptive adjustment element embedded in the entire online control loop: reducing the sampling frequency to save resources under low risk and increasing the sampling frequency to ensure response speed under high risk, ensuring that the control logic of Embodiments 2 to 5 maintains optimal performance under different operating conditions. Simultaneously, the sampling parameter adjustment history of this embodiment can also serve as a reference for control parameter diagnosis.

[0082] Example 7: In Examples 3 to 5, backflow events are triggered multiple times during operation. However, a single event often cannot reflect systemic problems. Only by recording and statistically analyzing the characteristics of multiple events can patterns be identified and the root cause located. Therefore, this example defines a unified criterion for judging backflow events, records characteristic parameters for each event, and calculates aggregate statistics after the statistical period ends.

[0083] See Figure 4 The backflow event recording module records a backflow event as one of the following situations within the statistical period: the backflow risk discrimination model determines that there is a risk and generates an anti-backflow control command; or the actual gate power is less than the first threshold for at least one sampling period.

[0084] For each reverse flow event, the following characteristic parameters are recorded: start time, end time, duration, minimum threshold power, whether the energy storage converter is in inverter state when the event occurs, energy storage converter response delay, energy storage converter power change rate, maximum allowable battery discharge power, battery state of charge, load power change, and photovoltaic power change.

[0085] The response delay of the energy storage converter is calculated as follows: from the moment the external EMS controller sends a power limiting command to the energy storage converter until the moment when the actual active power on the AC side of the energy storage converter begins to decrease and the decrease exceeds the dead zone threshold, the time elapsed.

[0086] After the statistical period ends, the statistical analysis module calculates the following aggregated statistics: total number of reverse current events, reverse current frequency, average energy storage converter response delay, average energy storage converter power change rate, proportion of non-inverter state events, proportion of short-term events, average maximum allowable battery discharge power, and proportion of events where the battery state of charge is below the lower limit.

[0087] The event logs and statistical results of this embodiment will serve as direct inputs to the fault diagnosis model of this embodiment. This embodiment is a data accumulation stage from online control to offline diagnosis, providing factual basis for subsequent root cause analysis and system optimization.

[0088] Example 8: Backflow can be caused by a variety of factors, such as sudden load changes, photovoltaic fluctuations, communication delays, insufficient regulation capacity of the energy storage converter, battery power limitations, or metering anomalies, which are difficult to quickly distinguish based on human experience alone. Therefore, this example establishes a hierarchical decision tree diagnostic model based on the event characteristics and statistics provided in Example 7, and judges the problem location layer by layer according to the order of elimination and causal chain.

[0089] See Figure 4 The fault diagnosis module uses a hierarchical decision tree, and the judgment process is as follows: The first layer determines whether a reverse current event occurs when the energy storage converter is in a non-inverter state. If so, it proceeds to the non-energy storage converter cause branch: if the photovoltaic power is greater than the load power during the event, it is determined to be independent photovoltaic backfeed; if the load power is negative, it is determined to be load feedback; otherwise, it is determined to be a metering sampling anomaly. If all events occur in the energy storage converter in the inverter state, it proceeds to the second layer.

[0090] The second layer determines whether the average response delay of the energy storage converter exceeds a preset response delay threshold. If so, it is determined to be a problem on the communication control side; otherwise, proceed to the third layer.

[0091] The third layer determines whether the power change rate of the energy storage converter is close to its maximum allowable value, and whether the power at the cut-off point remains below the first threshold during the event. If yes, proceed to the fourth layer; otherwise, proceed to the fifth layer.

[0092] The fourth layer: Determine the correlation between the event start time and the time of a sudden drop in load power or a sudden increase in photovoltaic power. If it is strongly correlated with a sudden drop in load power, it is determined that the change on the load side is too rapid; if it is correlated with a sudden increase in photovoltaic power, it is determined that the change on the photovoltaic side is too rapid.

[0093] Fifth layer: Determine if the event duration is extremely short and if there is an abnormal jump in the gate power. If so, determine that the metering sampling is abnormal; otherwise, determine that the control parameter settings are unreasonable.

[0094] Furthermore, based on battery state parameters, the following judgment branches can be added: if the maximum allowable discharge power of the battery is significantly lower than the actual power of the energy storage converter when multiple events occur, and the battery state of charge is close to the lower limit, then it is determined that the battery-side power capability is insufficient; if the battery temperature is abnormal when the event occurs, then it is determined to be a battery thermal management problem.

[0095] While outputting diagnostic results, the fault diagnosis module can also provide handling suggestions, such as adjusting control parameters, checking communication links, adding photovoltaic power limiting function, increasing the power change rate limit of energy storage converter, and optimizing battery state of charge management strategy.

[0096] The diagnostic results of this embodiment will be fed back into the integrated workflow to trigger adaptive adjustments to system parameters. For example, if the diagnosis is "unreasonable control parameter settings," the recovery threshold, window length, or prediction lead time can be automatically adjusted; if the diagnosis is "insufficient battery-side power capability," the battery SOC management threshold can be adjusted or maintenance prompts can be issued. Simultaneously, the problem location judgment output by this embodiment can also verify the effectiveness of the control logic in embodiments three to five, forming a closed-loop iteration of operation, recording, diagnosis, and optimization.

[0097] Example 9: The preceding examples addressed the issues of data acquisition and status recognition, feature extraction, risk assessment, battery coordination, recovery control, adaptive sampling, event recording and statistics, and fault diagnosis. However, these modules must be integrated into a cohesive whole and operate in a reasonable sequence to achieve maximum effectiveness. Therefore, this example unifies the online control and offline diagnostic components into a single workflow, clearly defining the execution order and data flow of each module.

[0098] See Figure 5 The system's overall workflow includes online control and offline diagnostics.

[0099] The online control section executes the following steps in each sampling cycle: The adaptive sampling control module determines the current sampling cycle and window length based on the risk margin index calculated in the previous sampling cycle, and sends the adjusted parameters to the gate power acquisition module and the time series analysis module; the gate power acquisition module acquires the current gate power and outputs it to the time series analysis module; the inverter status acquisition module acquires the energy storage converter status and outputs it to the control decision module; the battery status acquisition module acquires the battery's maximum allowable discharge power and state of charge and outputs it to the control decision module and the battery management coordination module; the time series analysis module extracts features and outputs them to the control decision module; the control decision module judges the reverse current risk. If the risk exists, it generates an anti-reverse current control command for the energy storage converter and sends it to the energy storage converter. At the same time, it triggers the battery management coordination module to send a battery-side power limit command to the battery management system and notifies the reverse current event recording module to start recording the event; if there is no risk, the recovery judgment module judges whether the recovery conditions are met. If they are met, the control decision module issues a recovery command and ends the event recording; the adaptive sampling control module updates the risk margin index.

[0100] The offline diagnostic module executes once per statistical time period: the reverse flow event recording module summarizes all events within that time period; the statistical analysis module calculates aggregate statistics; the fault diagnosis module outputs the problem location and suggestions through a hierarchical decision tree; and the diagnostic results are displayed through a human-machine interface or a remote monitoring platform, and can trigger adaptive adjustments to system parameters.

[0101] The online control section starts with the adaptive sampling in Example 6, sequentially calling the data acquisition in Example 1, the feature extraction in Example 2, and the risk judgment in Example 3. Based on the judgment results, it triggers the collaborative control in Example 4 or the recovery judgment in Example 5, and Example 7 continuously records events throughout the process. The offline diagnostic section summarizes the event data from Example 7 after each statistical time period, performs the hierarchical diagnosis in Example 8, and feeds the diagnostic results back to the parameter configuration of the online control section to achieve closed-loop optimization. Thus, Examples 1 to 8 are no longer isolated functional modules, but are organically connected into a complete system with perception, judgment, execution, recovery, learning, and optimization through the comprehensive workflow of this example.

[0102] Obviously, the embodiments described above are merely some, not all, embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort should fall within the scope of protection of the present invention.

Claims

1. An external EMS energy management system, wherein the external EMS is set independently of the energy storage converter and is communicatively connected to the gateway meter and the energy storage converter, characterized in that: It includes a gate power acquisition module, an inverter status acquisition module, a timing analysis module, a control decision module, a recovery judgment module, a battery status acquisition module, a battery management coordination module, an adaptive sampling control module, a reverse current event recording module, a statistical analysis module, and a fault diagnosis module; each module is located inside an external EMS and interacts with data and transmits commands through an internal data bus or software interface; The output of the gate power acquisition module is connected to the input of the timing analysis module; the output of the timing analysis module is connected to the input of the control decision module; the output of the inverter state acquisition module is connected to the input of the control decision module; the output of the battery state acquisition module is connected to the inputs of both the control decision module and the battery management coordination module; the output of the control decision module is connected to the input of the battery management coordination module; and the output of the battery management coordination module is connected to the battery management system. The gate power acquisition module is used to acquire time-series data of gate active power according to a preset sampling period; the inverter status acquisition module is used to acquire the operating status of the energy storage converter, at least distinguishing between inverter status and non-inverter status; the time-series analysis module is used to perform forward and backward trend analysis on the gate active power time-series data within the sliding time window and extract gate power change characteristics. The control decision module is used to determine whether there is a reverse current risk based on the threshold power change characteristics when the energy storage converter is in inverter mode, and to generate an anti-reverse current control command when the reverse current risk is determined to exist; the battery management coordination module is used to send a battery-side power limiting command; the recovery judgment module is used to release the anti-reverse current limit based on the threshold power change characteristics.

2. The external EMS energy management system according to claim 1, characterized in that: The gate power change characteristics include at least: current gate power, average window change rate, predicted gate power, and the number of sampling points within the window that are continuously below the first threshold. The average change rate of the window is the difference between the gate power at the end of the window and the gate power at the beginning of the window, divided by the window time span; the predicted gate power is the current gate power plus the product of the average change rate of the window and the preset lead time.

3. The external EMS energy management system according to claim 2, characterized in that: The control decision module determines that there is a risk of backflow when the following conditions are met simultaneously: The energy storage converter is in inverter mode; The average rate of change of the window is less than the first rate of change threshold; The current gate power is less than or equal to the first threshold, or the predicted gate power is less than or equal to the first threshold; The number of consecutive sampling points below the first threshold within the window reaches the preset number of times condition.

4. An external EMS energy management system according to claim 2, characterized in that: The recovery judgment module releases the anti-backflow restriction when the following conditions are met simultaneously: The current gate power is greater than the second threshold and continues for a preset number of sampling periods; The average rate of change of the window is greater than the second rate of change threshold; Wherein, the second threshold is greater than the first threshold.

5. An external EMS energy management system according to claim 1, characterized in that: The battery status acquisition module is used to obtain the maximum allowable discharge power of the battery from the battery management system; When judging the risk of reverse current, the control decision module also makes a joint judgment based on the maximum allowable discharge power of the battery. When the current actual discharge power of the energy storage converter exceeds the maximum allowable discharge power of the battery, it does not judge that there is a risk of reverse current or reduces the level of reverse current risk. While generating anti-reverse current control commands, the battery management coordination module sends battery-side power limiting commands to the battery management system, enabling the energy storage converter and the battery management system to work together.

6. An external EMS energy management system according to claim 5, characterized in that: The battery-side power limiting command includes at least one of the following: The upper limit of the battery's permissible discharge power will be temporarily adjusted to the target power limit value. When the power limit target value is zero or negative, the battery management system is instructed to switch the battery side to charging or standby mode. Adjust the upper limit of the battery discharge current so that the battery output power does not exceed the target power limit value.

7. An external EMS energy management system according to claim 6, characterized in that: The power limiting target value is the current threshold power minus the first threshold. When the power limit target value is positive, the energy storage converter is controlled to discharge at a power not exceeding the power limit target value; When the power limit target value is zero or negative, the energy storage converter is controlled to switch to standby or charging state.

8. An external EMS energy management system according to claim 2, characterized in that: The output of the adaptive sampling control module is connected to the gate power acquisition module and the time series analysis module, respectively; the adaptive sampling control module is used to dynamically adjust the sampling period and the sliding window length according to the risk margin index. The risk margin index is calculated based on the difference between the current gate power and the first threshold and the average rate of change of the window. When the risk margin index is larger, the sampling period is longer and the window length is smaller; when the risk margin index is smaller, the sampling period is shorter and the window length is larger.

9. An external EMS energy management system according to claim 2, characterized in that: The fault diagnosis module is used to determine the location of the backflow problem by using a hierarchical decision tree based on event feature parameters and statistical characteristics; the backflow event recording module is used to record a backflow event as a situation that meets the backflow risk judgment condition or the actual gate power is less than the first threshold for at least one sampling period within the statistical time period, and to record the event feature parameters. The statistical analysis module is used to aggregate and statistically analyze multiple reverse flow events within a statistical time period to obtain statistical characteristics.

10. An external EMS energy management system according to claim 9, characterized in that: The hierarchical decision tree includes: First layer: Determine whether there is a reverse current event in the non-inverter state of the energy storage converter. If so, determine whether it is photovoltaic independent backfeed, load feedback or metering sampling abnormality based on the relationship between photovoltaic power and load power. The second layer: Determine whether the average response delay of the energy storage converter is greater than the preset response delay threshold. If so, it is determined to be a problem on the communication control side. The third layer: Determine whether the power change rate of the energy storage converter is close to the maximum allowable value and whether the power at the cut-off point is still lower than the first threshold during the event. If so, proceed to the fourth layer. The fourth layer: Determine the correlation between the event and the sudden drop in load power or the sudden increase in photovoltaic power, and determine whether the change is too rapid on the load side or the photovoltaic side. Fifth layer: Determine whether the event duration is extremely short and whether there is an abnormal jump in the gate power. If so, it is determined to be an abnormal metering sampling; otherwise, it is determined to be an unreasonable control parameter setting.

Images