Energy storage system composite disturbance island detection and protection control method and system

Abstract

The application discloses a composite disturbance island detection and protection control method and system for an energy storage system. The method comprises the following steps: a dynamic frequency baseline is established and a frequency offset amount Δf is calculated when an energy storage converter PCS is in grid-connected operation; when |Δf| exceeds a first pre-judgment threshold δ1 and is less than a first disturbance threshold δ2, the PCS outputs a pure frequency bidirectional disturbance, and no additional reactive power disturbance is injected into the power grid; when |Δf| exceeds δ2 and is less than an island confirmation threshold δ x , the frequency disturbance direction is locked and a step reactive power disturbance is superimposed, the reactive power disturbance direction is consistent with the frequency disturbance direction, positive feedback acceleration is formed; when |Δf| exceeds δ x , islanding is confirmed and a protection action is performed. For a system configured with a static transfer switch STS, a decoupling protection mechanism and timing reset logic are used, that is, δ PCS > δ STS and T PCS ≥ 5×T STS , so that the timing competition between the PCS and the STS is avoided. The application realizes millisecond-level fast detection on the premise that the reactive power requirement of the grid-connected standard is met.

Description

Technical Field

[0001] This invention belongs to the field of energy storage system technology, specifically relating to a method and system for detecting and protecting against islanding caused by composite disturbances in energy storage systems. It is applicable to grid-connected and off-grid switching systems with static transfer switches (STS), and also applicable to pure grid-connected systems with only power conversion systems (PCS) configured. The latter performs protective shutdown after detecting islanding. Background Technology

[0002] Energy storage systems are an important component of smart grids and distributed energy resources, and their islanding detection function is an essential safety feature for grid-connected operation. Depending on the system configuration, the control objectives after islanding detection are divided into two categories (see...). Figure 1 , Figure 2 ): Table 1 Protection requirements for two different system configurations

[0003] Pure grid-connected systems are mainly designed for scenarios with no off-grid power supply needs, such as grid-side independent energy storage power stations and industrial and commercial energy storage that only participates in electricity market transactions.

[0004] Existing island detection methods are mainly divided into two categories: passive detection methods and active detection methods. Typical active detection methods include active frequency shift (AFD) and reactive power disturbance methods.

[0005] Active frequency shifting is easy to implement, but in practical applications, there is a trade-off between the detection dead zone (NDZ) and the disturbance parameter settings—if the disturbance is too small, the detection is slow; if the disturbance is too large, the harmonic distortion of the output current increases. To address this, the academic community has proposed several improvement schemes (such as variable frequency positive feedback coefficients), which have improved detection performance, but there is still a gap in meeting the millisecond-level switching requirements of energy storage systems.

[0006] The reactive power disturbance method disrupts the power balance in islanded conditions by injecting reactive power. This method has the following drawbacks: • GB / T 34133-2023 "Technical Specification for Testing Energy Storage Converters" requires that the grid-connected reactive power should approach zero during anti-islanding testing, which limits the use of reactive power disturbances in the prediction stage; • A single reactive power disturbance has a slow response in power matching scenarios; • Fixed amplitude square wave disturbances have a significant impact on local loads.

[0007] According to international anti-islanding test standards such as IEC 62116 and UL 1741, during testing, the RLC load must be adjusted so that the fundamental current flowing through the switch before grid disconnection is limited to below 1% to 2% of the rated output current, and the grid-connected reactive power approaches zero (or is in a balanced load state with Q=1.0). The Chinese national standard GB / T 34133 also adopts this test condition, requiring the grid-connected fundamental current to be less than 1% of the PCS rated current and the grid-connected reactive power to approach zero. This means: • Passive detection methods fail in power matching scenarios (detection blind zone). • Active detection methods must be performed only if they meet the standard requirements; • The prediction phase cannot rely on obvious changes in electrical quantities, nor can it use reactive disturbances.

[0008] Traditional existing technologies typically have the following shortcomings or urgent technical problems that need to be solved: (1) Detection blind zone and load matching problem: Traditional passive detection methods have a detection blind zone in power matching scenarios. Although the improved active frequency shift method and reactive power disturbance method can effectively overcome the blind zone and shorten the detection time, the existing improved schemes still have the following limitations: the detection blind zone of the active frequency shift method is contradictory to the frequency disturbance intensity control - if the disturbance is too small, the detection is slow, and if the disturbance is too large, the harmonic distortion will increase; the reactive power disturbance method is difficult to respond quickly under the initial conditions required by the standard (grid-connected reactive power is close to zero); although the bidirectional disturbance scheme can eliminate the blind zone, it lacks an active acceleration mechanism after the island is confirmed, and the detection time is usually evaluated in terms of power frequency cycle (about 20ms), which is difficult to meet the millisecond level requirements.

[0009] (2) Adaptability to power grid frequency fluctuations: Existing solutions mostly use fixed frequency thresholds for islanding detection, which can easily lead to false triggering when the power grid frequency fluctuates naturally. Although there is research on "dynamic protection thresholds", it mainly optimizes the protection response speed rather than achieving dynamic tracking of the frequency baseline.

[0010] (3) Insufficient detection speed: According to grid connection standards, grid-connected power supplies should disconnect from the grid within 2 seconds after islanding is detected. While existing active frequency shift and reactive power disturbance methods can meet this requirement, the detection time is typically in the hundreds of milliseconds to seconds. However, for energy storage systems equipped with static transfer switches (STS), the control objective after islanding is not shutdown, but rapid switching to off-grid mode. The switching time (including islanding detection) is generally required to be ≤100ms, with high-end requirements within 10ms, to ensure uninterrupted power supply to critical loads, far exceeding the standard requirement of a 2-second shutdown time. The response speed of existing detection methods is far from meeting this stringent requirement. Therefore, how to compress the islanding detection time from hundreds of milliseconds to milliseconds is the core technical problem to be solved. While improving the detection speed, it is also necessary to coordinate the triggering sequence with the STS to avoid premature shutdown of the PCS.

[0011] (4) Timing race problem in STS-equipped systems: In systems configured with STS, the islanding detection functions of PCS and STS may operate independently. If PCS detects islanding before STS and triggers protection shutdown, it will cause a power outage to the load, which contradicts the original intention of configuring STS. Therefore, the protection shutdown of PCS does not require the switching signal of STS, but is determined independently by its own threshold and time parameters; during normal switching, STS can still notify PCS to enter off-grid mode through communication.

[0012] In summary, existing technologies struggle to achieve millisecond-level islanding detection while meeting grid connection standards (grid-connected reactive power approaches zero during testing), and the timing competition between PCS and STS in STS systems remains unresolved. Summary of the Invention

[0013] Therefore, this invention provides a method and system for detecting and protecting against islanding in a complex disturbance in an energy storage system, a computer-readable storage medium, and an electronic device to solve the following technical problems: how to achieve millisecond-level fast islanding detection while meeting the reactive power requirements of grid connection standards, while solving the timing competition problem in systems with STS, and providing a unified islanding detection and control scheme applicable to both systems with and without STS.

[0014] A method for detecting and protecting against islanding caused by complex disturbances in an energy storage system includes the following steps: S1: When the energy storage converter PCS is in grid-connected operation, it monitors the grid frequency fmeas at its output terminal in real time, establishes a dynamic frequency baseline fbase, and calculates the frequency offset Δf. The frequency offset Δf is the difference between the measured grid frequency and the dynamic frequency baseline. The absolute value of the frequency offset |Δf| is used for threshold judgment, and the positive or negative sign of the frequency offset is used to judge the direction of disturbance. S2: When |Δf| reaches or exceeds the first prediction threshold δ1 and is less than the first disturbance threshold δ2, the energy storage converter outputs a bidirectional frequency disturbance. The bidirectional frequency disturbance periodically and alternately changes the disturbance direction, without changing the output current amplitude or injecting additional reactive power disturbance into the grid; where δ1≥0, when δ1=0, the system continuously outputs the bidirectional frequency disturbance under normal grid-connected operation, so that the prediction stage is always in an active state; S3: When |Δf| reaches or exceeds the first disturbance threshold δ2 and is less than the island confirmation threshold δ x When the frequency disturbance direction at the last moment in step S2 is locked to no longer alternate, the corresponding step is selected from multiple preset reactive power disturbance steps according to the magnitude of |Δf|, and the reactive power disturbance of the selected step is output; the direction of the reactive power disturbance is consistent with the locked frequency disturbance direction, and the larger the |Δf| is, the larger the reactive power disturbance amplitude corresponding to the selected step is. S4: When |Δf| reaches or exceeds the island confirmation threshold δ x And continuously confirm time T x Afterwards, confirm the island status and execute protective actions; Where δ1<δ2<δ x And δ1, δ2, δ x T x All are configurable parameters, T x Islanding detection time T for static transfer switch STS Islanding detection time T of the energy storage converter PCS .

[0015] Preferably, the dynamic frequency baseline is established through a first-order low-pass filter, and the filtering algorithm is: f base (n)=α·f base (n-1)+(1-α)·f meas (n), where f base (n) represents the dynamic frequency baseline of the nth sample; f base (n-1) is the dynamic frequency baseline of the (n-1)th sample; f meas (n) is the measured power grid frequency of the nth sampling; α is the filter coefficient, with a value range of 0.9~0.99.

[0016] Preferably, during the stepped perturbation stage, the frequency offset interval [δ2, δ] is... x According to the dividing points γ1, γ2, ..., γ m-1 Divide the system into m consecutive sub-intervals, where m ≥ 2. Each sub-interval corresponds to a first-level reactive power disturbance amplitude, with each level of reactive power disturbance amplitude being Q1, Q2, ..., Q... m Satisfying Q1 < Q2 < ... < Q mAll of these are configurable parameters. The reactive power disturbance Q is expressed as a piecewise constant function of Δf as follows:

[0017] Preferably, in step S3, the direction of the reactive power disturbance is consistent with the locked frequency disturbance direction, including: when the frequency disturbance direction is positive, outputting inductive reactive power to accelerate the frequency rise; when the frequency disturbance direction is negative, outputting capacitive reactive power to accelerate the frequency fall; the reactive power disturbance amplitudes at each stage are Q1, Q2, ..., Q... m The magnitude of the reactive power disturbance determines the degree to which it accelerates the grid frequency shift.

[0018] Preferably, the islanding detection and protection function of the energy storage converter is independent and completely decoupled from the islanding detection and protection function of the static transfer switch. Each is configured with a fixed detection threshold, confirmation time, and protection action: the energy storage converter has a fixed shutdown threshold δ. PCS Confirmation time T PCS and protection actions—perform protection shutdown and stop output; the STS is fixedly configured with a switching threshold δ STS Confirmation time T STS and protection actions—execute and switch off the network; and δ STS <δ PCS T STS <T PCS The island confirmation threshold δ x The protection action is determined as follows: For systems equipped with static transfer switches: δ x Take δ STS Confirmation time is T. STS The protection action is triggered by a static transfer switch and off-grid switching; the PCS independently monitors |Δf|, and when |Δf| reaches or exceeds δ PCS The timer starts at time T. If the timer continues for T seconds... PCS During this period, |Δf| never decreased to δ. PCS If the PCS then performs a backup protection shutdown, and |Δf| drops to δ during the timing period... PCS The timing is then reset; due to δ STS <δ PCS And T STS <T PCS After the static transfer switch switches normally, |Δf| returns to δ. PCS The following applies: the energy storage converter resets its timer without triggering shutdown; the backup protection shutdown of the energy storage converter is only triggered by itself when |Δf| continuously exceeds δ. PCS Reaching T PCSThe determination triggers the action, independent of the operating state or switching signal of the static transfer switch; it only occurs when the static transfer switch is not configured or its switching action is ineffective, causing |Δf| to continuously exceed δ. PCS Reaching T PCS At that time, the energy storage converter will perform backup protection shutdown; for systems without static transfer switches: δ x Take δ PCS Confirmation time is T. PCS The protection action is for the energy storage converter to perform a protective shutdown and stop output.

[0019] Preferably, T PCS ≥ 5×T STS This ensures that the PCS protection shutdown confirmation time is much longer than the STS switching time.

[0020] The present invention also provides a composite disturbance islanding detection and protection control system for an energy storage system, comprising: an energy storage unit, the energy storage unit including a battery system and an energy storage converter module; the energy storage converter module is configured to perform the above-described method.

[0021] Furthermore, the system also includes a static transfer switch module and / or a photovoltaic power generation unit. The static transfer switch module is connected to the energy storage converter module via a communication bus. The photovoltaic power generation unit includes a photovoltaic array and a maximum power point tracking (MPPT) controller.

[0022] In another aspect, the present invention provides a computer-readable storage medium and an electronic device, wherein a computer program stored thereon, when executed by a processor, implements all the steps of the above-described method.

[0023] The method and system of the present invention have at least the following beneficial effects: (1) Good standard compliance: The first stage of pure frequency disturbance does not change the output power and does not inject additional reactive disturbance components into the power grid, meeting the requirements of domestic and foreign anti-islanding test standards such as GB / T 34133-2023, IEC 62116, and UL 1741.

[0024] (2) Fast detection speed: In the second stage of composite disturbance, frequency disturbance and reactive disturbance are superimposed in the same direction to form positive feedback, which significantly accelerates frequency shift. The detection time can be compressed from hundreds of milliseconds to milliseconds, meeting the fast switching requirements of STS system.

[0025] (3) Strong grid adaptability: The dynamic frequency baseline can track the natural fluctuations of the power grid and avoid false triggering; the first prediction threshold δ1 can be 0, covering two working modes: continuous disturbance and threshold triggering.

[0026] (4) Rapid detection avoids baseline drift risk: Since the total switching time of the STS system is no more than 10ms, which is much smaller than the response time constant of the dynamic frequency baseline, the Δf attenuation problem caused by the baseline "lagging follow" is effectively avoided.

[0027] (5) Small load impact: Frequency disturbance does not change the current amplitude, and reactive power disturbance is increased step by step to avoid sudden disturbance from impacting the local load.

[0028] (6) Wide range of applications: The unified detection framework is applicable to grid-connected and off-grid switching systems with STS and pure grid-connected systems; the configurable threshold parameters are adapted to grid-connection standards in different countries and regions.

[0029] (7) Low implementation cost: Pure software control strategy, no hardware modification required; pure grid-connected system does not require additional STS configuration.

[0030] (8) Step-adaptive acceleration: In the step-disturbance stage, the reactive disturbance amplitude increases adaptively with the frequency offset |Δf|, that is, the larger |Δf| is, the larger the reactive disturbance amplitude corresponding to the selected step is; thus, the frequency offset accumulates at a faster rate in the high-step interval, so that the detection process automatically accelerates convergence and further shortens the island confirmation time.

[0031] (9) Timing race avoidance: By setting δ PCS >δ STS And T PCS ≥ 5×T STS Combined with the mechanism of automatic reset of PCS timer when frequency recovers, it ensures that STS in the STS system completes the switching first, and after the frequency recovers, PCS timer is reset without triggering shutdown, thus avoiding PCS shutdown causing power failure to the load.

[0032] (10) Backup protection mechanism: When the STS handover fails, the PCS will detect a frequency offset that continues to exceed δ. PCS Reaching T PCS Automatic backup shutdown provides dual security.

[0033] (11) The system is configurable for all threshold parameters (δ1, δ2, δ...). STS δ PCS T PCS T STS T freq Both the parameters and the step parameters can be flexibly configured according to the target market standards, making them highly universal. Attached Figure Description

[0034] Figure 1 This is a system architecture diagram of an embodiment of the present invention (Configuration A: a complete system with STS, including optional photovoltaics); Figure 2 This is a system architecture diagram of an embodiment of the present invention (Configuration B: Pure grid-connected system); Figure 3 This is a schematic diagram of the composite perturbation threshold layering system according to an embodiment of the present invention; Figure 4 This is a flowchart of the composite disturbance control method according to an embodiment of the present invention; Figure 5 This is a diagram showing the correspondence between frequency offset and reactive power disturbance step in an embodiment of the present invention. Figure 6 This is a schematic diagram of the bidirectional frequency disturbance waveform in the first stage of an embodiment of the present invention; Figure 7 This is a flowchart of island detection and protection (including timing race avoidance mechanism) according to an embodiment of the present invention. Detailed Implementation

[0035] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of protection of the present invention.

[0036] This invention provides a method for detecting and protecting against islanding caused by complex disturbances in an energy storage system, comprising the following steps: S1: When the energy storage converter PCS is in grid-connected operation, it monitors the grid frequency fmeas at its output terminal in real time, establishes a dynamic frequency baseline fbase, and calculates the frequency offset Δf. The frequency offset Δf is the difference between the measured grid frequency and the dynamic frequency baseline. The absolute value of the frequency offset |Δf| is used for threshold judgment, and the positive or negative sign of the frequency offset is used to judge the direction of disturbance. S2: When |Δf| reaches or exceeds the first prediction threshold δ1 and is less than the first disturbance threshold δ2, the energy storage converter outputs a bidirectional frequency disturbance. The bidirectional frequency disturbance periodically and alternately changes the disturbance direction, without changing the output current amplitude or injecting additional reactive power disturbance into the grid; where δ1≥0, when δ1=0, the system continuously outputs the bidirectional frequency disturbance under normal grid-connected operation, so that the prediction stage is always in an active state; S3: When |Δf| reaches or exceeds the first disturbance threshold δ2 and is less than the island confirmation threshold δ x When the frequency disturbance direction at the last moment in step S2 is locked to no longer alternate, the corresponding step is selected from multiple preset reactive power disturbance steps according to the magnitude of |Δf|, and the reactive power disturbance of the selected step is output; the direction of the reactive power disturbance is consistent with the locked frequency disturbance direction, and the larger the |Δf| is, the larger the reactive power disturbance amplitude corresponding to the selected step is. S4: When |Δf| reaches or exceeds the island confirmation threshold δ xAnd continuously confirm time T x Afterwards, confirm the island status and execute protective actions; Where δ1<δ2<δ x And δ1, δ2, δ x T x All are configurable parameters, T x Islanding detection time T for static transfer switch STS Islanding detection time T of the energy storage converter PCS .

[0037] Another aspect of this invention provides a composite disturbance islanding detection and protection control system for an energy storage system, comprising: an energy storage unit, the energy storage unit including a battery system and an energy storage converter module; the energy storage converter module is configured to perform the method described above.

[0038] The energy storage system in this embodiment of the invention supports two configuration methods, each suitable for different application scenarios.

[0039] Configuration A: On-grid / off-grid handover system with STS. For example... Figure 1 As shown, the configuration includes: Energy storage units mainly include battery systems and energy storage converters (PCS); Static transfer switch (STS module) is used to quickly switch to off-grid power supply mode when the grid is interrupted; Local load; Communication bus (CAN bus, Ethernet, etc.) is used for communication between PCS and STS; Optional photovoltaic power generation unit (PV photovoltaic array and maximum power point tracking MPPT controller).

[0040] In this configuration, the PCS executes the aforementioned composite disturbance islanding detection and protection control method according to this embodiment of the invention during grid-connected operation. When an islanding state is detected, the STS responds when the frequency offset reaches δ. STS Time delay T STS Triggering and switching off-grid mode, and notifying the PCS to switch to off-grid operating mode via the communication bus. The PCS simultaneously and independently monitors the frequency offset: when |Δf| reaches or exceeds δ... PCS The timer starts at time T. If the timer continues for T seconds... PCS During this period, |Δf| never decreased to δ. PCS The following will trigger a backup shutdown; if |Δf| drops to δ during the timing period. PCS Then, the timer will be reset.

[0041] Configuration B: Pure grid-connected system (without STS). For example... Figure 2 As shown, this configuration is suitable for pure grid-connected systems with only PCS configured, including: Energy storage units mainly include battery systems and PCS; Local load (connected via grid connection point); Optional photovoltaic power generation unit.

[0042] This configuration does not include the STS module.

[0043] In this configuration, the PCS performs the same composite disturbance islanding detection method. When an islanding state is detected, the PCS responds when the frequency offset reaches δ. PCS And continue T PCS Then, it performs a protective shutdown, stops outputting to the power grid, and disconnects from the grid within 2 seconds as required by the standard.

[0044] To avoid false triggering caused by natural fluctuations in the power grid, embodiments of this invention introduce a dynamic frequency baseline. For example... Figure 3 As shown, the dynamic frequency baseline is established through a first-order low-pass filter, and its mathematical expression is: f base (n)=α·f base (n-1)+(1-α)·f meas (n) Where: f base (n) represents the dynamic frequency baseline of the nth sample; f base (n-1) is the dynamic frequency baseline of the (n-1)th sample; f meas (n) is the measured power grid frequency of the nth sampling; α is the filter coefficient, with a value range of 0.9~0.99 (or can be adjusted according to actual needs).

[0045] Frequency offset is defined as: Δf=f meas -f base The dynamic frequency baseline slowly tracks the natural fluctuations of the power grid frequency, separating normal frequency changes from abnormal offsets and reducing the risk of false triggering. Upon initial power-on, the current measured frequency is used as the initial value of the dynamic frequency baseline, i.e., f. base (0)=f meas (0). During islanding detection, because the embodiments of the present invention employ a fast switching mechanism (the total switching time for STS systems typically does not exceed 10ms) or a stepped positive feedback disturbance (continuously accelerating frequency offset), the time from the start of the frequency offset to reaching the threshold is much smaller than the response time constant of the dynamic baseline, for example, differing by at least one order of magnitude. Therefore, the baseline's hysteresis effect will not affect the reliability of islanding judgment. For purely grid-connected systems, if the detection time is long (e.g., close to seconds), a shorter T can be configured. PCS Alternatively, a fixed-frequency baseline can be used to further avoid the Δf attenuation problem that may be caused by baseline "lag following".

[0046] Based on the relationship between the frequency offset |Δf| and a preset threshold, the system operates in the following three stages, corresponding to steps S2, S3, and S4 as described above, and as follows: Figure 3 , Figure 4 As shown in Table 2.

[0047] Table 2. Three stages of the disturbance and protection process

[0048] Where δ x δ is determined based on system configuration. STS (STS switching threshold for STS systems) or δ PCS (PCS shutdown threshold for a pure grid-connected system).

[0049] The hierarchical relationship of the threshold parameters is: δ1 < δ2 < δ x δ1 is the first prediction threshold, which can be 0, meaning the system can continuously output bidirectional frequency disturbances in normal conditions; δ2 is the first disturbance threshold, which is the critical point for entering the stepped disturbance stage.

[0050] In stage 2 (step disturbance stage), i.e., step S3, the reactive power disturbance is not a fixed value, but rather the disturbance amplitude increases stepwise as the frequency offset increases, forming a step-increasing disturbance sequence, such as... Figure 5 As shown.

[0051] Suppose the frequency offset |Δf| is within the stepped perturbation interval [δ2, γ]. m The interval is divided into m consecutive sub-intervals (m≥2), corresponding to m levels of reactive power disturbance. The step boundary points satisfy δ2<γ1<γ2<…<γ m-1 <γ m The direction of reactive power disturbance is consistent with the direction of Δf offset, that is, when Δf>0, the output is inductive reactive power (Q>0), and when Δf<0, the output is capacitive reactive power (Q<0).

[0052] The magnitude of the reactive disturbance, Q, is expressed as a piecewise constant function of Δf, as shown in Table 3 below.

[0053] Table 3. Explanation of the classification of stepped reactive power disturbances

[0054] The piecewise constant function formula for reactive disturbance Q as Δf is expressed as:

[0055] Among them, the reactive power disturbance amplitudes at each level are Q1, Q2, ..., Q... m The condition 0 < Q1 < Q2 < ... < Q is satisfied. m All of these are configurable parameters.

[0056] When |Δf|≥γ m At that time, the maximum amplitude |Q| = Q is maintained. m The number of steps remains unchanged. The number of steps can be set according to system requirements (e.g., 2-5 steps), and the amplitude of each step can be adjusted according to PCS capacity, load characteristics, and standard requirements. The maximum disturbance amplitude is limited by PCS capacity and power quality requirements.

[0057] The technical effects of the stepped design are: (1) smooth transition - avoid the impact of sudden disturbances on the local load; (2) positive feedback acceleration - the greater the frequency offset, the stronger the disturbance, further accelerating the frequency offset; (3) adaptability - the number of steps and the amplitude of each step can be flexibly configured according to the system capacity and standard requirements.

[0058] The following is an example of a 3-level stepped reactive power disturbance, where Pn is the rated power of the PCS: Table 4—Examples of Level 3 Stepped Reactive Power Disturbance

[0059] The above step boundaries and amplitudes are just examples. The number of steps, the boundaries of each level, and the amplitudes can be flexibly configured according to system capacity, standard requirements, and application scenarios.

[0060] In this embodiment of the invention, the frequency disturbance and reactive power disturbance are directionally consistent, forming a positive feedback acceleration effect.

[0061] In Phase 1 (prediction phase), the PCS outputs bidirectional frequency disturbances: during a positive disturbance, the current frequency is slightly higher than the grid frequency, causing the frequency to rise; during a negative disturbance, the current frequency is slightly lower than the grid frequency, causing the frequency to fall; the two directions alternate periodically, with each direction lasting for a preset time T. freq The detection criterion is: calculate the frequency offset |Δf|, and if it exceeds the threshold δ2, it is confirmed that the power grid may be disconnected.

[0062] Upon entering Phase 2 (Step-by-Step Disturbance Phase): The frequency disturbance direction is locked, no longer alternating, maintaining the frequency disturbance direction from the last moment of Phase 1; reactive power disturbances are superimposed, with the reactive power disturbance direction consistent with the locked frequency disturbance direction. Specifically, this includes: when the frequency disturbance direction is positive, outputting inductive reactive power to accelerate frequency increase; when the frequency disturbance direction is negative, outputting capacitive reactive power to accelerate frequency decrease; the reactive power disturbance amplitudes at each stage are Q1, Q2, ..., Q... m The magnitude of the reactive power disturbance determines the degree to which it accelerates the grid frequency shift. The directional correspondence is shown in Table 5 below: Table 5. Disturbance direction and disturbance effect

[0063] Frequency disturbances and reactive power disturbances superimpose in the same direction, forming positive feedback, which significantly accelerates the frequency shift towards the islanding confirmation threshold δ. STS or δ PCS .like Figure 6 The diagram shown is a schematic of the bidirectional frequency disturbance waveform in the first stage.

[0064] In a system with STS, there are two levels of control requirements: normal handover—STS quickly switches over after detecting islanding and notifies PCS to enter off-network working mode; protective shutdown—when the system is not configured with STS or STS handover fails, PCS should automatically perform protective shutdown.

[0065] This invention embodiment completely decouples the protection shutdown function of the PCS from the normal switching of the STS: both the PCS and STS are configured with fixed detection thresholds, confirmation times, and protection actions: the PCS of the energy storage converter is configured with a fixed shutdown threshold δ. PCS Confirmation time T PCS and protection actions—perform protection shutdown and stop output; the STS is fixedly configured with a switching threshold δ STS Confirmation time T STS and protection actions—execute and switch off the network; and δ STS <δ PCS T STS <T PCS The PCS's protective shutdown does not depend on any signal from the STS, but only on its own detected |Δf| and delay T. PCS Independent judgment. Parameter settings ensure that the PCS will not be accidentally triggered to shut down during normal STS switching.

[0066] Table 6 STS and PCS Parameter Setting Requirements

[0067] The control logic is as follows: (1) STS: Independent detection |Δf|≥δ STS After a delay T STS The handover is triggered and the network connection is switched off after a period of time (usually no more than 10ms).

[0068] (2) PCS: Independent detection |Δf|≥δ PCS When |Δf| reaches or exceeds δ PCS The timer starts at time T. If the timer continues for T seconds... PCS During this period, |Δf| never decreased to δ. PCS The following will trigger a backup protection shutdown (stop output); if |Δf| drops to δ during the timing period. PCS Then, the timer is reset. This protective shutdown does not depend on any signal from the STS; it is entirely based on its own parameters.

[0069] (3) In a system with STS: due to δ PCS >δ STS And T PCS ≥ 5×T STS The STS will inevitably complete the switchover first. After the switchover, the load will be powered by energy storage, the frequency offset will return to normal, and |Δf| will not continue to reach δ. PCS Therefore, the protection shutdown conditions of the PCS are not met, and it will not malfunction.

[0070] (4) In a pure grid-connected system (without STS) or in an STS failure scenario: |Δf| will continue to rise and exceed δ PCS And the duration reaches T PCS PCS automatically performs shutdown protection.

[0071] like Figure 7 The diagram shown is a flowchart of island detection and protection (including timing race avoidance mechanism).

[0072] This invention further configures the functional relationship between control logic and parameters, rather than specific parameter values. All threshold and time parameters are configurable to adapt to grid connection standards and system configurations in different countries and regions.

[0073] Table 7 Configurable parameters applicable to different needs

[0074] Configuration methods include: configuration via host computer software, configuration via field configuration tools, or remote configuration via communication bus.

[0075] Combination Figure 4 The complete control flow of the method according to the embodiments of the present invention is described in detail below: Step S10: The system is connected to the grid and the PCS monitors its own output voltage frequency in real time. It updates the dynamic frequency baseline through a first-order low-pass filter and calculates the frequency offset Δf.

[0076] Step S20: When |Δf| exceeds the first predicted threshold δ1 (δ1≥0, can be 0), initiate Phase 1: Output bidirectional frequency perturbation. The frequency perturbation direction is periodically and alternately changed, with each direction lasting a preset time T. freq Detect whether |Δf| exceeds the first perturbation threshold δ2.

[0077] Step S30: When |Δf| reaches or exceeds the first disturbance threshold δ2, it is confirmed that the power grid may be disconnected, and the process enters stage 2: stepped disturbance stage. The current frequency disturbance direction is locked to no longer alternate; the corresponding step is selected from multiple preset reactive power disturbance steps based on the magnitude of |Δf|; the reactive power disturbance of the selected step is output, with the reactive power disturbance direction consistent with the locked frequency disturbance direction; when |Δf| increases to the boundary point of the next step, a larger reactive power disturbance amplitude is switched to.

[0078] Step S40: Perform Phase 3 protection according to system configuration.

[0079] The specific execution method of step S40 varies depending on the system configuration and can be divided into two types: Configuration A (System with STS): Normal STS handover – when |Δf| reaches or exceeds the STS handover threshold δ STS At that time, after a delay of T STS (Typically no more than 10ms) Triggering and switching off-network, and notifying the PCS to switch to off-network working mode via communication. PCS backup protection shutdown (decoupled and independent) - The PCS continuously monitors Δf, and when |Δf| reaches or exceeds the PCS shutdown threshold δ PCS (δ) PCS >δ STS The timer starts at time T. If the timer continues for T seconds... PCS (T) PCS ≥ 5×T STS During this period, |Δf| never decreased to δ. PCS The following will automatically trigger a backup protection shutdown (stop output); if |Δf| drops to δ during the timing period. PCS The timer will then be reset. This backup protection shutdown does not depend on the STS signal and will not be triggered during normal STS switching; it only takes effect automatically when the STS switching fails.

[0080] Configuration B (pure grid-connected system, without STS): When |Δf| reaches or exceeds the PCS shutdown threshold δ PCS And continue T PCS When the PCS confirms the island, it performs a protective shutdown and stops output.

[0081] After the protection action is completed, the PCS exits the disturbance mode.

[0082] The following specific embodiments illustrate the parameter configuration and working process of the present invention in different application scenarios.

[0083] Example 1: China's power grid scenario This embodiment 1 corresponds to the application scenario of the Chinese power grid. According to GB / T 34120-2023, the normal operating range of the power grid is 48.5Hz~50.5Hz. The system adopts configuration A (grid-connected and off-grid switching system with STS), which includes a battery system, a PCS module, an STS module, and a local load.

[0084] Table 8 Example values ​​of parameters in Example 1

[0085] Operating Process: When the system is connected to the grid, the PCS establishes a dynamic frequency baseline with α = 0.95 and continuously outputs a bidirectional frequency disturbance with an amplitude of 0.2 Hz (δ1 = 0.03 Hz, approximately continuous disturbance). When the grid is de-energized, the frequency offset |Δf| gradually increases. When |Δf| exceeds 0.2 Hz (δ2), the PCS locks the disturbance direction and superimposes the first-order reactive power disturbance (2%P). n When |Δf| reaches 0.5Hz, switch to the second step (4%P). n ), accelerating frequency shift. When |Δf| reaches 1.5Hz (δ STS And lasted for 4ms (T) STS After this, the STS is triggered and a handover to the offline network begins, and the PCS enters offline operating mode. If the STS handover fails, |Δf| will continue to rise to 2.0Hz (δ). PCS And lasts for 100ms (T) PCS ), PCS performs backup protection shutdown.

[0086] Example 2: European EN 50549 scenario This Example 2 corresponds to the European power grid application scenario, and according to the EN 50549 standard, the power grid operating range is ±2Hz. The system adopts Configuration B (pure grid-connected system, without STS), which includes a battery system, a PCS module, and local loads.

[0087] Table 9 Example values ​​of parameters in Example 2

[0088] Operating Procedure: During grid-connected operation, the PCS establishes a dynamic frequency baseline with α = 0.9. When |Δf| exceeds 0.05Hz (δ1), bidirectional frequency disturbance is initiated. When |Δf| exceeds 0.2Hz (δ2), the PCS locks the disturbance direction and sequentially superimposes the first step (1%P). n ), second tier (3%P) n ), third step (5%P) n Reactive disturbances accelerate frequency shift. Due to the absence of STS, when |Δf| reaches 2.5Hz (δ... PCS And lasts for 200ms (T) PCSAfter that, the PCS performs a protective shutdown and stops output. This embodiment uses a 3-level step, which, compared to the 2-level step in embodiment 1, provides a finer disturbance gradient over a wider frequency offset range, adapting to the larger operating range of ±2Hz in the European power grid.

[0089] Example 3: North American IEEE 1547 scenario This embodiment 3 corresponds to a North American power grid application scenario, based on the IEEE 1547 standard. The system adopts configuration A (grid-connected / off-grid switching system with STS), including a battery system, a PCS module, an STS module, and a local load. This scenario is characterized by a small allowable range of grid frequency fluctuations and high requirements for detection sensitivity.

[0090] Table 10 Example values ​​of parameters in Example 3

[0091] Operating Procedure: When the system is connected to the grid, the PCS establishes a dynamic frequency baseline with α = 0.95. When |Δf| exceeds 0.1Hz (δ1), a bidirectional frequency disturbance with an amplitude of 0.3Hz is initiated. When |Δf| exceeds 0.3Hz (δ2), the PCS locks the disturbance direction and superimposes the first-step reactive power disturbance (2%P). n When |Δf| exceeds 0.8Hz, switch to the second step (5%P). n ), accelerating frequency shift. When |Δf| reaches 1.0Hz (δ STS ) and lasts for 10ms (T) STS After this, the STS is triggered and a handover to the offline network begins, and the PCS enters offline operating mode. If the STS handover fails, |Δf| will continue to rise to 1.5Hz (δ). PCS ) and last for 100ms (T) PCS The PCS performs backup protection shutdown. In this embodiment, compared to Embodiment 1, δ1 and δ2 are larger (0.1Hz and 0.3Hz vs 0.03Hz and 0.2Hz) to accommodate the higher frequency disturbance amplitude requirements of the North American power grid; δ STS and δ PCS The smaller values ​​(1.0Hz and 1.5Hz vs 1.5Hz and 2.0Hz) reflect the stricter restrictions imposed by IEEE 1547 on frequency anomalies.

[0092] The above embodiments are merely illustrative of the feasibility of the present invention and are not intended to limit the scope of protection. The three embodiments correspond to three application scenarios: the Chinese power grid (GB / T 34120-2023), the European power grid (EN 50549), and the North American power grid (IEEE 1547), employing different parameter configurations and system architectures (configuration A or configuration B), demonstrating the universality of the parameter system of the present invention. The number of steps, the boundary points of each level, and the amplitude can all be flexibly configured according to system capacity, standard requirements, and application scenarios. The present invention uses the same detection framework and parameter system to measure δ... x T PCS T STS With the configuration of parameters such as STS, it can seamlessly adapt to systems with and without STS, and has strong versatility.

[0093] The above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for detecting and protecting against islanding caused by complex disturbances in an energy storage system, characterized in that, Includes the following steps: S1: When the energy storage converter is in grid-connected operation, it monitors the grid frequency f at its output terminal in real time. meas Establish dynamic frequency baseline f base Calculate the frequency offset Δf, where Δf is the difference between the measured grid frequency and the dynamic frequency baseline. Take the absolute value of the frequency offset |Δf| for threshold judgment, and use the positive or negative sign of the frequency offset for disturbance direction judgment. S2: When |Δf| reaches or exceeds the first prediction threshold δ1 and is less than the first disturbance threshold δ2, the energy storage converter outputs a bidirectional frequency disturbance. The bidirectional frequency disturbance periodically and alternately changes the disturbance direction, without changing the output current amplitude or injecting additional reactive power disturbance into the grid; where δ1≥0, when δ1=0, the system continuously outputs the bidirectional frequency disturbance under normal grid-connected operation, so that the prediction stage is always in an active state; S3: When |Δf| reaches or exceeds the first disturbance threshold δ2 and is less than the island confirmation threshold δ x When the frequency disturbance direction at the last moment in step S2 is locked to no longer alternate, the corresponding step is selected from multiple preset reactive power disturbance steps according to the magnitude of |Δf|, and the reactive power disturbance of the selected step is output; the direction of the reactive power disturbance is consistent with the locked frequency disturbance direction, and the larger the |Δf| is, the larger the reactive power disturbance amplitude corresponding to the selected step is. S4: When |Δf| reaches or exceeds the island confirmation threshold δ x And continuously confirm time T x Afterwards, confirm the island status and execute protective actions; Where δ1<δ2<δ x And δ1, δ2, δ x T x All are configurable parameters, T x Islanding detection time T for static transfer switch STS Islanding detection time T of the energy storage converter PCS .

2. The method according to claim 1, characterized in that, The dynamic frequency baseline is established through a first-order low-pass filter, and the filtering algorithm is: f base (n)=α·f base (n-1)+(1-α)·f meas (n), where f base (n) represents the dynamic frequency baseline of the nth sample; f base (n-1) is the dynamic frequency baseline of the (n-1)th sample; f meas (n) is the measured power grid frequency of the nth sampling; α is the filter coefficient, with a value range of 0.9~0.

99.

3. The method according to claim 1, characterized in that, During the stepped perturbation phase, the frequency offset interval [δ2, δ] is... x According to the dividing points γ1, γ2, ..., γ m-1 Divide the system into m consecutive sub-intervals, where m ≥ 2. Each sub-interval corresponds to a first-level reactive power disturbance amplitude, with each level of reactive power disturbance amplitude being Q1, Q2, ..., Q... m Satisfying Q1 < Q2 < ... < Q m All of these are configurable parameters. The reactive power disturbance Q is a piecewise constant function of Δf, expressed as: 。 4. The method according to claim 3, characterized in that, In step S3, the direction of the reactive power disturbance is consistent with the locked frequency disturbance direction, including: when the frequency disturbance direction is positive, outputting inductive reactive power to accelerate the frequency rise; when the frequency disturbance direction is negative, outputting capacitive reactive power to accelerate the frequency fall; the amplitudes of the reactive power disturbances at each stage are Q1, Q2, ..., Q... m The magnitude of the reactive power disturbance determines the degree to which it accelerates the grid frequency shift.

5. The method according to claim 1, characterized in that, The islanding detection and protection function of the energy storage converter is independent and completely decoupled from the islanding detection and protection function of the static transfer switch. Each is configured with a fixed detection threshold, confirmation time, and protection action. The energy storage converter has a fixed shutdown threshold δ. PCS Confirmation time T PCS and protection actions—perform protection shutdown and stop output; the static transfer switch is fixedly configured with a switching threshold δ. STS Confirmation time T STS and protection actions—execute and switch off the network; and δ STS <δ PCS T STS <T PCS The island confirmation threshold δ x The protection action is determined as follows: For systems equipped with static transfer switches: δ x Take δ STS Confirmation time is T. STS The protection action is triggered by a static transfer switch and off-grid switching; the energy storage converter independently monitors |Δf|, and when |Δf| reaches or exceeds δ PCS The timer starts at time T. If the timer continues for T seconds... PCS During this period, |Δf| never decreased to δ. PCS If the energy storage converter then performs backup protection shutdown, and |Δf| drops to δ during the timing period... PCS The timing is then reset; due to δ STS <δ PCS And T STS <T PCS After the static transfer switch switches normally, |Δf| returns to δ. PCS The following applies: the energy storage converter resets its timer without triggering shutdown; the backup protection shutdown of the energy storage converter is only triggered by itself when |Δf| continuously exceeds δ. PCS Reaching T PCS The determination triggers the action, independent of the operating state or switching signal of the static transfer switch; it only occurs when the static transfer switch is not configured or its switching action is ineffective, causing |Δf| to continuously exceed δ. PCS Reaching T PCS At that time, the energy storage converter will perform backup protection shutdown; for systems without static transfer switches: δ x Take δ PCS Confirmation time is T. PCS The protection action is for the energy storage converter to perform a protective shutdown and stop output.

6. The method according to claim 5, characterized in that, T PCS ≥ 5×T STS This ensures that the shutdown confirmation time of the energy storage converter protection is much longer than the switching time of the static transfer switch.

7. A composite disturbance islanding detection and protection control system for an energy storage system, characterized in that, include: Energy storage unit, the energy storage unit including a battery system and an energy storage converter module; The energy storage converter module is configured to perform the method according to any one of claims 1-6.

8. The system according to claim 7, characterized in that, It also includes a static transfer switch module and / or a photovoltaic power generation unit. The static transfer switch module is connected to the energy storage converter module via a communication bus. The photovoltaic power generation unit includes a photovoltaic array and a maximum power point tracking (MPPT) controller.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1-6.

10. An electronic device, characterized in that, It includes a processor and a memory, wherein the memory stores a computer program that, when executed by the processor, implements the method as described in any one of claims 1-6.

Images