Island detection method based on zero non-detection area direct current micro-grid local parameters

Abstract

The invention provides an island detection method based on local parameters of a zero non-detection area direct current micro-grid. The island detection method comprises the following steps of performing tiny working point disturbance on a photovoltaic system; the condition of a real-time digital simulator is considered; setting a voltage limit value and a threshold value of an accumulated voltage error; detailed results are discussed for different events and network configurations, including zero power mismatch and non-zero power mismatch. According to the invention, by introducing a local parameter-based accumulated voltage error island detection method, the problem of a non-detection area of a traditional method is solved, and island detection with zero non-detection area, low cost, quick response and minimum influence on electric energy quality is innovatively realized. According to the invention, the system realization is simplified, the detection reliability is enhanced, and a new solution is provided for island detection of the DC micro-grid.

Description

Technical Field

[0001] This invention relates to the field of power grid detection technology, and in particular to an islanding detection method based on local parameters of a DC microgrid in a zero-detection zone. Background Technology

[0002] In DC microgrids, islanding can lead to power system instability, making accurate islanding detection crucial. Traditional islanding detection techniques mainly include communication-based and local parameter-based methods. Communication-based methods rely on external signal transmission, requiring additional hardware and communication links, increasing cost and system complexity. Furthermore, communication failures can cause detection delays or failures. Local parameter-based methods (such as voltage and current variations) typically suffer from non-detection zones (NDZ), making them ineffective at detecting islanding under specific load conditions. Active detection methods avoid NDZ by injecting disturbances, but this significantly impacts power quality, especially in DC microgrids. To overcome these limitations, a novel islanding detection method is urgently needed, offering advantages such as zero non-detection zone, no communication required, and minimal impact on power quality, enabling reliable and rapid detection of islanding effects under various operating conditions. Summary of the Invention

[0003] To address the problems existing in the prior art, this invention proposes an islanding detection method based on local parameters of a DC microgrid in the zero non-detection zone, which can detect islands in a DC microgrid within a maximum duration of 0.11s.

[0004] To achieve the above objectives, the present invention adopts the following technical solution:

[0005] An islanding detection method based on local parameters of a DC microgrid in the zero-detection zone includes the following steps:

[0006] S1. Perform a small operating point disturbance on the photovoltaic system;

[0007] S2. Consider the case of a real-time digital simulator;

[0008] S3, Set the voltage limit value and the threshold value for cumulative voltage error;

[0009] S4. Discuss detailed results for different events and network configurations, including zero power mismatch and non-zero power mismatch.

[0010] Preferably, in step S1, the photovoltaic unit is integrated with a DC-DC converter to regulate the voltage loop and generate the maximum possible power by operating the unit at the maximum power point, employing a DC-DC boost converter, wherein the DC link voltage V DClink The relationship with the terminal voltage of the photovoltaic unit is as follows:

[0011]

[0012] In the formula, D is the duty cycle;

[0013] In grid-connected mode, the voltage of the DC bus is regulated by the grid, while in islanded mode, the voltage of the DC bus is not regulated, defined as...

[0014]

[0015] In the formula V DClink P is the DC link voltage during islanding. PV R represents the actual output power of the photovoltaic unit. equi The equivalent load resistance during islanding is determined by turning on the relevant switch in the boost converter for 0.1s to confirm islanding, during which the duty cycle D is temporarily disturbed.

[0016] Preferably, in step S1, the disturbance attempts to shift the operating point of the photovoltaic unit towards the open-circuit voltage V of the photovoltaic array. OC Therefore, regardless of the operating mode of the DC microgrid, the output power of the photovoltaic array will drop instantaneously, re-tracking the violation point of the maximum power point in 0.17 seconds. Under the same scenario, the coaxial supply voltage exhibits different behavior. In islanded mode, the coaxial supply voltage deviates significantly, while in grid-connected mode, due to grid voltage support, the coaxial supply voltage is less affected. Because the coaxial supply voltage changes significantly, monitoring the cumulative voltage error of 0.1 seconds helps to perceive the operating mode of the DC microgrid. The formula for calculating the cumulative voltage error (CVE) is...

[0017]

[0018] Among them, VE i To define a voltage reference (V) DC =300V voltage error), n is the number of samples used for accumulation (1667 data samples taken for 0.1s at 60μs sampling).

[0019] Preferably, in step S2, modeling is performed in a real-time digital simulator. There is an interleaved bidirectional AC / DC converter at the grid end, and a bidirectional DC / DC converter in the battery energy storage system. The AC / DC bidirectional converter operates in DC / AC mode and is defined based on the DC bus voltage. The DC / DC synchronous buck-boost converter operates in buck or boost mode. The photovoltaic unit is integrated through the DC / DC boost converter to achieve incremental conductance MPP tracking. The schedulable proton exchange membrane FC is integrated through the DC / DC boost converter to provide better energy support during islanding.

[0020] Preferably, in step S3, the maximum observed change in the voltage limit value Vdel during steady-state operation of the DC microgrid is approximately 0.4V; the minimum observed change in the voltage limit value Vdel during steady-state operation is due to available grid voltage support and effective bus voltage regulation, approaching the nominal bus voltage of the converter integrated source; conversely, significant changes are observed in the voltage limit value Vdel depending on the nature of transient events, up to 160V for high-frequency radio communication and as low as 10V for islanding events.

[0021] Preferably, in step S4, the zero-power mismatch islanding event: the effective load of the DC system can be modeled as constant impedance and power loads. Due to the existence of negative resistance transient behavior, the increase in the proportion of constant power loads compared to constant impedance loads makes islanding detection easier. Therefore, even a small disturbance may escalate to a significant level. In addition, the worst islanding condition is zero power mismatch, with only constant impedance loads. Under the condition of zero power mismatch, a zero-power mismatch islanding event is created in 0.14s, and different proportions of constant impedance and power loads are successfully detected within 0.11s. Even with zero constant power loads, the zero-power mismatch condition can be met. If the CVE of the coaxial power supply bus increases significantly, this is due to the considerable voltage deviation caused by the small converter switching and the inability to obtain grid voltage support during islanding. As the penetration of constant power loads increases, the islanding detection time decreases from 0.11s to 0.1s. The worst islanding zero-power mismatch constant impedance load is 12kW, the constant power load is 0kW, and the non-detection zone is zero.

[0022] Preferred non-zero power mismatch islanding events: Islanding events are created in a DC microgrid where the 15kW and 9kW loads have considerable non-zero power mismatch conditions during islanding. Voltage limits and cumulative voltage error changes are analyzed to effectively detect non-zero power mismatch islanding events of 0.09s and 0.09s.

[0023] Compared with the prior art, the present invention has the following beneficial effects:

[0024] This invention solves the non-detection zone problem of traditional methods by introducing a cumulative voltage error islanding detection method based on local parameters. It innovatively achieves islanding detection with zero non-detection zone, low cost, fast response, and minimal impact on power quality. This application not only simplifies system implementation but also enhances detection reliability, providing a new solution for islanding detection in DC microgrids.

[0025] This application distinguishes between islanded and non-islanded events by monitoring the cumulative change in coaxial supply voltage; it uses transient disturbances to adjust the operating point of the photovoltaic system, achieving zero non-detection zone (Zero NDZ) islanding detection with minimal impact on power quality; it relies solely on the single local parameter of coaxial supply voltage for detection, eliminating the need for communication links or external signal transmission, thus reducing system costs and improving reliability; it features rapid detection (≤0.11 seconds) and high adaptability under various network topologies (such as ring and radial structures) and load conditions (including zero power mismatch and non-zero power mismatch), making it suitable for efficient islanding detection in DC microgrids and possessing broad application prospects. Attached Figure Description

[0026] Figure 1 The PV characteristic curve;

[0027] Figure 2 Design diagram for islanding detection method considering DC microgrids. Detailed Implementation

[0028] The technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0029] This embodiment proposes an islanding detection method based on local parameters of a DC microgrid in the zero-detection zone, which detects islands in a DC microgrid within a maximum duration of 0.11s, including the following steps:

[0030] S1. Perform a small perturbation on the operating point of the photovoltaic (PV) system.

[0031] First, the PV unit is integrated with a DC-DC converter to regulate the voltage loop and generate the maximum possible power by operating the unit at the maximum power point (MPP), such as... Figure 1 As shown. A DC-DC boost converter is used, where the DC link voltage (V) DClink ) and PV terminal voltage (V PV The relationship is

[0032]

[0033] In the formula, D represents the duty cycle. In grid-connected mode, the DC bus voltage is regulated by the grid, while in islanded mode, the DC bus voltage is not regulated and can be defined as...

[0034]

[0035] In the formula V DClink P is the DC link voltage during islanding. PV R represents the actual output power of the PV.equi This represents the equivalent load resistance during islanding. Islanding is confirmed by turning on the relevant switch in the boost converter for 0.1s; in the proposed technique, D is temporarily disturbed. For example... Figure 1 As shown, this disturbance attempts to shift the PV operating point toward the open-circuit voltage (V) of the PV array. OC Therefore, regardless of the DC microgrid's operating mode, the PV array's output power will drop instantaneously, re-tracking the MPP violation point (within 0.17s). However, the POC (coaxial power supply) voltage behaves differently under the same scenario. In islanded mode, the POC voltage deviates significantly, while in grid-connected mode, the POC voltage is less affected due to grid voltage support. Because of the significant POC voltage variation, monitoring the CVE for 0.1s helps to perceive the DC microgrid's operating mode. The CVE calculation formula is...

[0036]

[0037] Among them, VE i To define a voltage reference (V) DC =300V voltage error), n is the number of samples used for accumulation (1667 data points sampled for 0.1s at 60μs). The disturbance in D is instantaneous, and its impact on power quality is negligible. Furthermore, only when the superimposed absolute POC voltage (V...)... del =|V t -V t-1 Disturbance is triggered only when the change exceeds a system-independent threshold, where V t It is the POC voltage sampled instantaneously at time t. This avoids unnecessary triggering of the proposed interference and limits the impact on power quality. CVE vs. V del This method is used to reliably distinguish between isolated and non-isolated transient events. This application analyzes the proposed scheme for different types of events. Based on this analysis, an appropriate system-independent threshold V is selected. del This threshold can be further lowered to improve the reliability of the proposed technique. Further RTDS-based results demonstrate the reliability of the proposed island detection technique due to its single-parameter dependence.

[0038] S2. Consider the case of RTDS (Real-Time Digital Simulator).

[0039] The proposed islanding detection technology was tested on a networked DC microgrid, such as... Figure 2As shown, two microgrids (MG1 and MG2) represent two villages equipped with energy storage devices such as PV, fuel cells (FC), and battery energy storage systems (BESS) to meet local needs. The test system can be modeled in either a ring or radial configuration by properly operating the circuit breakers. The system is modeled in RTDS with an interleaved bidirectional AC / DC converter at the grid end and a bidirectional DC / DC converter at the BESS end. The bidirectional AC / DC converter can operate in DC / AC mode and has a defined reference based on the DC bus voltage. Similarly, the DC / DC synchronous buck-boost converter can operate in either buck or boost mode. The PV units are integrated via a DC / DC boost converter for incremental conductance MPP tracking. The schedulable proton exchange membrane FC is integrated via a DC / DC boost converter to provide better energy support during islanding.

[0040] S3. Set the thresholds for Vdel and CVE.

[0041] The proposed technique requires intelligent threshold setting to improve system reliability while maintaining a practical level of system safety. Therefore, the proposed scheme was thoroughly analyzed for several transient and steady-state events. During steady-state operation of the DC microgrid, the maximum observed change in Vdel was approximately 0.4V. The minimum change in Vdel observed during steady-state operation was due to available grid voltage support and effective bus voltage regulation, approaching the nominal bus voltage of the converter integrated source. Conversely, significant changes in Vdel were observed depending on the nature of transient events, reaching up to bb0160V for HRF (High Frequency Radio Communication) and as low as ≈10V for islanding events. Therefore, to ensure reliable detection of islanding events, we initially set the Vdel threshold to +5 for transient event detection. It should be noted that the Vdel threshold can be increased or decreased, depending on the required trade-off between sensitivity to transient events and disturbance injection. Furthermore, to distinguish these transients from islanding and non-islanding events, the MPP point was temporarily perturbed, and the corresponding change in CVE at the POC was observed. During non-islanding events, the maximum recorded CVE is 120. Considering an additional 25% margin, the CVE threshold is chosen to be 120 × 1.25 = 150. To avoid unexpected tripping, a disturbance is injected after a 0.02s delay in Vdel deviation detection, ensuring the overall islanding detection time is less than or equal to 0.11s.

[0042] S4. Discuss detailed results for different events and network configurations, including ZPM (Zero Power Mismatch) and non-ZPM.

[0043] ZPM Islanding Event: The effective load of a DC system can be modeled as constant impedance and power loads. Due to the potential for negative resistance transient behavior, increasing the proportion of constant power loads relative to constant impedance loads makes islanding detection easier. Therefore, even small disturbances can escalate to a significant level. Furthermore, the worst islanding condition is ZPM, with only constant impedance loads. In microgrid 1, under the ZPM condition, a ZPM islanding event is created in 0.14s, and different proportions of constant impedance and power loads are successfully detected within 0.11s, even with zero constant power loads, the ZPM condition is met. If the CVE of the POC bus increases significantly, it is due to the considerable voltage deviation caused by small converter switching and the lack of grid voltage support during islanding. As the penetration of constant power loads increases, the islanding detection time decreases (from 0.11s to 0.1s). Therefore, this scheme effectively detects the worst-case islanding ZPM condition (12kW constant impedance load, 0kW constant power load), with zero NDZ.

[0044] Non-ZPM isolated events: To further evaluate the behavior of the proposed solution under non-ZPM conditions, deliberately in Figure 2 Islanding events were created in a DC microgrid (MG1) where a 15kW (+25% PV capacity) and a 9kW (-25% PV capacity) load exhibited significant non-ZPM (Pmdel) conditions during islanding. Analysis of Vdel and CVE variations in POC1 revealed that the scheme effectively detected non-ZPM islanding events at 0.09s and 0.09s.

[0045] The islanding detection method proposed in this embodiment, based on the calculation of cumulative voltage error (CVE), distinguishes between islanded and non-islanded events by monitoring the cumulative change in POC voltage. It uses transient disturbances to adjust the operating point of the photovoltaic system, achieving zero-NDZ islanding detection with minimal impact on power quality. Detection is performed using only the single local parameter of POC voltage, eliminating the need for communication links or external signal transmission, thus reducing system costs and improving reliability. It exhibits rapid detection (≤0.11 seconds) and high adaptability under various network topologies (such as ring and radial structures) and load conditions (including ZPM and non-ZPM). This scheme is suitable for efficient islanding detection in DC microgrids and has broad application prospects.

[0046] By introducing a local parameter-based cumulative voltage error (CVE) islanding detection method, the non-detection zone problem of traditional methods is solved, innovatively achieving islanding detection with zero non-detection zone, low cost, fast response, and minimal impact on power quality. This technology not only simplifies system implementation but also enhances detection reliability, providing a new solution for islanding detection in DC microgrids.

[0047] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. An islanding detection method based on local parameters of a DC microgrid in a zero-detection zone, characterized in that, Includes the following steps: S1. Perform a small operating point disturbance on the photovoltaic system; S2. Consider the case of a real-time digital simulator; S3, Set the voltage limit value and the threshold value for cumulative voltage error; S4. Discuss detailed results for different events and network configurations, including zero power mismatch and non-zero power mismatch.

2. The islanding detection method based on local parameters of a DC microgrid in a zero-detection zone according to claim 1, characterized in that, In step S1, the photovoltaic unit is integrated with a DC-DC converter to regulate the voltage loop and generate the maximum possible power by operating the unit at the maximum power point. A DC-DC boost converter is used, wherein the DC link voltage V DClink The relationship with the terminal voltage of the photovoltaic unit is as follows: In the formula, D is the duty cycle; In grid-connected mode, the voltage of the DC bus is regulated by the grid, while in islanded mode, the voltage of the DC bus is not regulated, defined as... In the formula V DClink P is the DC link voltage during islanding. PV R represents the actual output power of the photovoltaic unit. equi The equivalent load resistance during islanding is determined by turning on the relevant switch in the boost converter for 0.1s to confirm islanding, during which the duty cycle D is temporarily disturbed.

3. The islanding detection method based on local parameters of a DC microgrid in a zero-detection zone according to claim 2, characterized in that, In step S1, the disturbance attempts to shift the operating point of the photovoltaic unit towards the open-circuit voltage V of the photovoltaic array. OC Therefore, regardless of the operating mode of the DC microgrid, the output power of the photovoltaic array will drop instantaneously, re-tracking the violation point of the maximum power point in 0.17 seconds. Under the same scenario, the coaxial supply voltage exhibits different behavior. In islanded mode, the coaxial supply voltage deviates significantly, while in grid-connected mode, due to grid voltage support, the coaxial supply voltage is less affected. Because the coaxial supply voltage changes significantly, monitoring the cumulative voltage error of 0.1 seconds helps to perceive the operating mode of the DC microgrid. The formula for calculating the cumulative voltage error (CVE) is... Among them, VE i To define a voltage reference (V) DC =300V voltage error), n is the number of samples used for accumulation (1667 data samples taken for 0.1s at 60μs sampling).

4. The islanding detection method based on local parameters of a DC microgrid in a zero-detection zone according to claim 1, characterized in that, In step S2, a model is created in a real-time digital simulator. There is an interleaved bidirectional AC / DC converter at the grid end, and a bidirectional DC / DC converter in the battery energy storage system. The AC / DC bidirectional converter operates in DC / AC mode and is defined based on the DC bus voltage. The DC / DC synchronous buck-boost converter operates in buck or boost mode. The photovoltaic unit is integrated through the DC / DC boost converter to achieve incremental conductance MPP tracking. The schedulable proton exchange membrane FC is integrated through the DC / DC boost converter to provide better energy support during islanding.

5. The islanding detection method based on local parameters of a DC microgrid in a zero-detection zone according to claim 1, characterized in that, In step S3, during steady-state operation of the DC microgrid, the maximum observed change in the voltage limit Vdel is approximately 0.4V; the minimum observed change in the voltage limit Vdel during steady-state operation is due to available grid voltage support and effective bus voltage regulation, approaching the nominal bus voltage of the converter integrated source; conversely, significant changes are observed in the voltage limit Vdel depending on the nature of transient events, reaching up to 160V for high-frequency radio communication and as low as 10V for islanding events.

6. The islanding detection method based on local parameters of a DC microgrid in a zero-detection zone according to claim 1, characterized in that, In step S4, zero-power mismatch islanding events: The effective load of the DC system can be modeled as constant impedance and power loads. Due to the existence of negative resistance transient behavior, the increase in the proportion of constant power loads compared to constant impedance loads makes islanding detection easier. Therefore, even a small disturbance may escalate to a significant level. Furthermore, the worst islanding condition is zero power mismatch, with only constant impedance loads. Under the condition of zero power mismatch, a zero-power mismatch islanding event is created in 0.14s, and different proportions of constant impedance and power loads are successfully detected within 0.11s. Even with zero constant power loads, the zero-power mismatch condition can be met. If the CVE of the coaxial power supply bus increases significantly, this is due to the considerable voltage deviation caused by the small converter switching and the inability to obtain grid voltage support during islanding. As the penetration of constant power loads increases, the islanding detection time decreases from 0.11s to 0.1s. The worst-case islanding zero-power mismatch constant impedance load is 12kW, the constant power load is 0kW, and the non-detection zone is zero.

7. The islanding detection method based on local parameters of a DC microgrid in a zero-detection zone according to claim 6, characterized in that, Non-zero power mismatch islanding events: Islanding events are created in a DC microgrid where 15kW and 9kW loads have considerable non-zero power mismatch conditions during islanding. Voltage limits and cumulative voltage error changes are analyzed to effectively detect non-zero power mismatch islanding events at 0.09s and 0.09s.

Images