A current proportioning and current limiting control method and controller for multi-inverter grid-connected system

Abstract

The application discloses a current proportional distribution and current limiting control method and controller of multiple grid-connected inverters, wherein when a complete network topology is unknown, initial virtual impedance of each grid-connected inverter is calculated according to a single current share corresponding to a reference virtual resistance, a reference virtual inductance and a target current distribution weight; when the topology is known, virtual impedance scaling values of each grid-connected inverter are calculated according to the topology, so that a deviation between a prediction and the target current share is minimized, and the initial virtual impedance is calculated accordingly; and then the initial virtual impedance is corrected according to actual current shares obtained through simulation or field test, so as to improve current proportional distribution accuracy of large-scale grid-connected inverters. Virtual impedance voltage drops generated by output current vectors of each grid-connected inverter are sent to an input end of a controller of the corresponding inverter, so that the corresponding inverter meets a virtual impedance-free contracted Jacobian matrix, and thus current proportional distribution and voltage synchronization can be realized under normal and load mutation working conditions.

Description

Technical Field

[0001] This invention belongs to the field of grid-connected inverter control technology in smart grids, and more specifically, relates to a method and controller for proportional current distribution and current limiting control of multiple grid-connected inverters. Background Technology

[0002] In renewable energy power plants, energy storage power stations, microgrids, and weak grid grid-connected systems, a large number of power electronic inverters undertake the tasks of voltage formation, frequency support, and power supply. Grid-mounted inverters can generate AC voltage even without strong grid support and are important equipment for future smart grids and high-proportion renewable energy power systems.

[0003] When multiple grid-connected inverters are in operation, the following engineering objectives must be met simultaneously: First, the output voltage and frequency of each inverter should be synchronized; second, under normal operating conditions, the current should be distributed proportionally according to capacity, dispatch, or available power; third, during faults such as PCC voltage drop, ABC grounding fault, or AB two-phase short circuit, overcurrent should be limited, and voltage, frequency, and current distribution should be restored after the fault is cleared.

[0004] However, after voltage synchronization of multiple grid-connected inverters, the current share ratios flowing through the corresponding loads may not meet actual engineering requirements. Therefore, it is necessary to readjust the current share of each grid-connected inverter. Existing technologies typically require the design of a power loop for current share allocation, but this method cannot achieve precise current share allocation under unplanned load conditions, resulting in poor accuracy in the proportional current distribution of the grid-connected inverters.

[0005] On the other hand, conventional virtual impedance current distribution, adaptive virtual impedance current limiting, and network synchronization control are typically designed separately. Virtual impedance alters voltage commands and branch equivalent impedances; especially during faults when virtual impedance is changed online by a PI controller, if its effect directly enters the vector field of the synchronization controller, the original contraction criterion may need to be re-analyzed with the virtual impedance. Therefore, it is necessary to ensure that the additional virtual impedance, while achieving current regulation, does not enter the contraction vector field used for voltage synchronization determination. This balances the proportional current distribution and voltage synchronization under normal operation and load surges, while simultaneously achieving overcurrent limiting and voltage synchronization during faults. Summary of the Invention

[0006] To address the aforementioned deficiencies or improvement needs of existing technologies, this invention provides a method and controller for proportional current distribution and current limiting control of multiple grid-connected inverters, thereby solving the technical problem of poor accuracy in the current proportional distribution method of existing grid-connected inverters under unplanned load conditions.

[0007] To achieve the above objectives, according to a first aspect of the present invention, a method for proportional current distribution and current limiting control of multiple grid-connected inverters is provided, comprising: S1, determine the target current allocation weight and target current share of each grid inverter; S2, when the complete network topology is unknown, calculate the initial virtual impedance of each network inverter based on the reference virtual resistance and reference virtual inductance corresponding to the single current share and the target current allocation weight of each network inverter. When the complete network topology is known, the initial virtual impedance scaling value corresponding to each grid inverter is set, and the virtual impedance scaling value corresponding to each grid inverter is adjusted with the goal of minimizing the deviation between the predicted current share and the target current share of each grid inverter under the current virtual impedance scaling value. Based on the optimized virtual impedance scaling value corresponding to each grid inverter obtained by adjustment, the initial virtual impedance corresponding to the corresponding grid inverter is calculated. The predicted current share of any grid-connected inverter under the current virtual impedance scaling condition is determined based on the following method: Based on the node admittance matrix corresponding to the current network topology, the node admittance equation and the equivalent branch current equation of each grid inverter are solved simultaneously to obtain the equivalent branch current vector corresponding to the current network topology. Based on the equivalent branch current of each grid inverter contained in the equivalent branch current vector, the predicted current share of each grid inverter under the current virtual impedance scaling value is calculated. S3. Based on the initial virtual impedance corresponding to each grid inverter, simulation or testing is performed to obtain the actual current share corresponding to each grid inverter. Based on the actual current share and target current share corresponding to each grid inverter, the initial virtual impedance corresponding to the corresponding grid inverter is optimized to obtain the calibrated virtual impedance corresponding to each grid inverter that can achieve its target current share. S4, the controller of the k-th grid inverter obtains the virtual impedance drop based on the output current vector and its calibrated virtual impedance; the controller stores the virtual impedance drop, uses it to generate the voltage command of the k-th grid inverter, and sends the virtual impedance drop to the corresponding input terminal of the controller to correct the internal voltage state change rate.

[0008] Based on the current proportional distribution and current limiting control method for the above-mentioned multiple grid-connected inverters, the node admittance equation is: ; The equivalent branch current equation for the kth grid-connected inverter is: ; Where Y is the node admittance matrix corresponding to the current network topology, and v is the access node voltage vector corresponding to the current network topology. Let v be the voltage at the access node corresponding to the k-th grid-connected inverter. This represents the equivalent branch current vector corresponding to the current network topology. for The equivalent branch current corresponding to the k-th grid inverter The output voltage of the k-th grid-connected inverter; , and Here, ω0 represents the filter resistor and filter inductor of the k-th grid-connected inverter, and ω0 is the rated angular frequency. , and These are the reference virtual resistance and reference virtual inductance corresponding to a single current share, respectively. This is the current virtual impedance scaling value corresponding to the kth grid-connected inverter; After obtaining the equivalent branch current vector corresponding to the current network topology, the predicted current share of the k-th grid inverter under the current virtual impedance scaling condition is calculated based on the following formula. : ; Where N represents the total number of grid-connected inverters.

[0009] Based on the current proportional distribution and current limiting control method for multiple grid-connected inverters described above, the virtual impedance scaling value corresponding to any grid-connected inverter is adjusted according to the following formula: ; in, and These represent the virtual impedance scaling values ​​before and after adjustment for the k-th grid-connected inverter, respectively; η is the relaxation factor, and the sat function is a saturation function whose output is limited to its minimum value. and maximum value between, The target current share for the k-th grid-connected inverter. For the k-th grid inverter in Predicted current share under certain conditions.

[0010] Based on the current proportional distribution and current limiting control method for multiple grid-connected inverters, when the complete network topology is unknown, the initial virtual impedance corresponding to each grid-connected inverter is calculated based on the following formula: ; ; in, and These are the reference virtual resistance and reference virtual inductance corresponding to a single current share, respectively. Assign weights to the target current of the k-th grid-connected inverter. and These are the initial virtual resistance and initial virtual inductance corresponding to the k-th grid inverter, respectively. The initial virtual impedance corresponding to the k-th grid inverter is composed of the initial virtual resistance and initial virtual inductance.

[0011] Based on the current proportional distribution and current limiting control method for multiple grid-connected inverters described above, after obtaining the initial virtual impedance when the complete network topology is unknown or known, the initial virtual impedance corresponding to each grid-connected inverter is optimized based on the following formula: ; ; in, and These are the calibration virtual resistor and calibration virtual inductor corresponding to the k-th grid-connected inverter, respectively. The calibration virtual impedance corresponding to the k-th grid-connected inverter is composed of the calibration virtual resistor and calibration virtual inductor. γ is the correction index. and The initial virtual resistance and initial virtual inductance are obtained by decomposing the initial virtual impedance corresponding to the k-th grid-connected inverter. This represents the actual current share corresponding to the k-th grid-connected inverter. This represents the target current share of the k-th grid-connected inverter.

[0012] Based on the aforementioned method for proportional current distribution and current limiting control of multiple grid-connected inverters, the method further includes: When a phase-to-phase short circuit or asymmetrical voltage drop occurs between phases A and B, the current limiting input of each grid inverter is obtained; wherein, the current limiting input of any grid inverter is obtained by normalizing the maximum value of the three-phase output current of the grid inverter. If the current limiting input of any grid inverter exceeds the preset current limiting target, the virtual impedance corresponding to the grid inverter will be increased by the PI controller based on its rated virtual impedance to limit the current limiting input of the grid inverter.

[0013] Based on the current proportional distribution and current limiting control method for multiple grid-connected inverters, the PI controller increases the virtual impedance corresponding to the grid-connected inverter based on the following formula: ; ; ; in, The current limiting input for the k-th grid-connected inverter at time t. For the preset rate limiting target, Let q be the current limiting error at time t. I,k (t) represents the state variable of the k-th grid-connected inverter corresponding to the integral element of the PI controller, and Δt is the controller sampling period. The virtual impedance corresponding to the k-th grid-connected inverter obtained after adjustment by the PI controller. The virtual impedance corresponding to the k-th grid-connected inverter is... and These are the proportional and integral coefficients of the PI controller, respectively.

[0014] Based on the aforementioned method for proportional current distribution and current limiting control of multiple grid-connected inverters, the step of increasing the virtual impedance corresponding to the grid-connected inverter via a PI controller to limit the current limiting input of the grid-connected inverter further includes: Real-time acquisition of current limiting input from each grid-connected inverter; If the current limiting input of any grid inverter is lower than the release threshold within the preset release time, the virtual impedance corresponding to that grid inverter will be reset to its corresponding calibrated virtual impedance.

[0015] Based on the aforementioned method for proportional current distribution and current limiting control of multiple grid-connected inverters, the method further includes: The virtual impedance corresponding to each grid inverter is incorporated into the voltage command of that grid inverter, so that the voltage command of that grid inverter satisfies: ; in, Let be the reference voltage that the k-th grid-connected inverter actually outputs to the voltage control loop at time t. Let be the synchronization voltage indicated by the synchronization voltage command at time t. This refers to either the calibrated virtual impedance corresponding to the k-th grid-connected inverter or the virtual impedance corresponding to the k-th grid-connected inverter after adjustment by the PI controller. Let be the output current vector of the k-th grid-connected inverter in the stationary αβ coordinate system.

[0016] According to a second aspect of the invention, a current proportional distribution and current limiting controller for multiple grid-connected inverters is provided for performing the method described in the first aspect.

[0017] According to a third aspect of the present invention, an electronic device is provided, comprising: a computer-readable storage medium and a processor; The computer-readable storage medium is used to store executable instructions; The processor is configured to read executable instructions stored in the computer-readable storage medium and execute the method as described in the first aspect.

[0018] According to a third aspect of the invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to perform the method as described in the first aspect.

[0019] According to a fourth aspect of the invention, a computer program product is provided, comprising a computer program or instructions that, when executed by a processor, implement the method described in the first aspect.

[0020] In summary, compared with the prior art, the above-described technical solutions conceived by this invention can achieve the following beneficial effects: When the complete network topology is unknown, the initial virtual impedance of each grid inverter is calculated based on the reference virtual resistance and reference virtual inductance corresponding to a single current share and the target current allocation weight of each grid inverter. The feasible first-round proportional current allocation can be obtained by inversely proportionalizing the virtual impedance with the target current allocation weight. When the complete network topology is known, the initial virtual impedance scaling value of each grid inverter is set, and the virtual impedance scaling value of each grid inverter is adjusted with the goal of minimizing the deviation between the predicted current share and the target current share of each grid inverter under the current virtual impedance scaling value. Based on the optimized virtual impedance scaling value of each grid inverter obtained by adjustment, the initial virtual impedance of the corresponding grid inverter is calculated as an alternative path for black-box parameter tuning. On this basis, through simulation or field test calibration, the initial virtual impedance of the corresponding grid inverter is optimized based on the actual current share and the target current share of each grid inverter. This can reduce the residual allocation error in the complete site model and improve the accuracy of proportional current allocation in large-scale grid inverter scenarios.

[0021] Furthermore, during faults such as PCC voltage drop, ABC grounding fault, and AB unbalanced short circuit, by acquiring the current limiting input of each grid inverter, when the current limiting input of any grid inverter exceeds the preset current limiting target, the PI controller increases the virtual impedance corresponding to that grid inverter based on its calibrated virtual impedance to limit the current limiting input of that grid inverter. This can quickly increase the equivalent impedance, limit overcurrent, and avoid sudden surges after fault clearance. Simultaneously, the controller sends the virtual impedance voltage drop to the corresponding input terminal to correct the internal voltage state change rate, ensuring that the Jacobian matrix of the dynamic equation of the output voltage of the grid inverter controlled by the controller does not contain virtual impedance and satisfies the contraction property. This allows the grid inverter to achieve proportional current distribution under normal operating conditions and maintain output voltage synchronization during fault current limiting. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of the control structure for multiple grid-connected inverters connected to an AC network via virtual impedance, as provided in an embodiment of the present invention.

[0023] Figure 2 This is a bar chart showing the current allocation error of random set A provided in an embodiment of the present invention.

[0024] Figure 3 The following are voltage drop waveforms for AB unbalanced short circuit and ABC ground fault on the PCC side provided in the embodiments of the present invention, wherein (a) corresponds to AB unbalanced short circuit fault and (b) corresponds to ABC ground fault.

[0025] Figure 4 The adaptive virtual impedance current limiting and synchronous recovery response diagram under ABC grounding fault provided in the embodiment of the present invention shows (a) the maximum phase current limiting input Ipu,k and 1.5pu current limiting threshold of all 33 grid-connected inverters, (b) the virtual impedance change after PI regulation, (c) the output voltage amplitude recovery, and (d) the frequency recovery.

[0026] Figure 5 The above is a response diagram of adaptive virtual impedance current limiting and synchronous recovery under AB unbalanced short circuit fault provided in an embodiment of the present invention, wherein (a) shows the maximum phase current limiting process, (b) shows the virtual impedance adjustment process, (c) shows the voltage amplitude recovery, and (d) shows the frequency recovery. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0028] This invention provides a method for proportional current distribution and current limiting control of multiple grid-connected inverters, comprising: S1, determine the target current allocation weight and target current share of each grid inverter; S2, when the complete network topology is unknown, calculate the initial virtual impedance of each network inverter based on the reference virtual resistance and reference virtual inductance corresponding to the single current share and the target current allocation weight of each network inverter. When the complete network topology is known, the initial virtual impedance scaling value corresponding to each grid inverter is set, and the virtual impedance scaling value corresponding to each grid inverter is adjusted with the goal of minimizing the deviation between the predicted current share and the target current share of each grid inverter under the current virtual impedance scaling value. Based on the optimized virtual impedance scaling value corresponding to each grid inverter obtained by adjustment, the initial virtual impedance corresponding to the corresponding grid inverter is calculated. The predicted current share of any grid-connected inverter under the current virtual impedance scaling condition is determined based on the following method: Based on the node admittance matrix corresponding to the current network topology, the node admittance equation and the equivalent branch current equation of each grid inverter are solved simultaneously to obtain the equivalent branch current vector corresponding to the current network topology. Based on the equivalent branch current of each grid inverter contained in the equivalent branch current vector, the predicted current share of each grid inverter under the current virtual impedance scaling value is calculated. S3. Based on the initial virtual impedance corresponding to each grid inverter, simulation or testing is performed to obtain the actual current share corresponding to each grid inverter. Based on the actual current share and target current share corresponding to each grid inverter, the initial virtual impedance corresponding to the corresponding grid inverter is optimized to obtain the calibrated virtual impedance corresponding to each grid inverter that can achieve its target current share. S4, the controller of the k-th grid inverter obtains the virtual impedance drop based on the output current vector and its calibrated virtual impedance; the controller stores the virtual impedance drop, uses it to generate the voltage command of the k-th grid inverter, and sends the virtual impedance drop to the corresponding input terminal of the controller to correct the internal voltage state change rate.

[0029] Here, for N grid-connected inverters operating in parallel or at a field station, the rated current I of each grid-connected inverter is defined. rated,k Filter resistor r f,k Filter inductor L f,k Point of Common Coupling (PCC) voltage v PCC Three-phase output current i a,k i b,k i c,k and the target current allocation weight ρ k Among them, grid-connected inverters include voltage source grid-connected inverters in new energy power plants, wind farms, photovoltaic power plants, energy storage power plants, or microgrids. Target current allocation weight ρ k This represents the relative current carrying capacity or dispatch weight of the k-th grid-connected inverter. For example, a larger ρ can be used for a grid-connected inverter with a larger capacity and higher available power. kSimultaneously, based on the target current allocation weight of each grid inverter, the target current share s of each grid inverter in the entire network is calculated. k Here, the target current share s k It can be defined as: .

[0030] To achieve proportional current distribution of grid-connected inverters so that their actual current share in the network is sufficiently close to their target current share, this embodiment of the invention introduces virtual impedance into the branches of the grid-connected inverters to adjust their local equivalent impedance, thereby achieving the corresponding current distribution of the grid-connected inverters. For example... Figure 1 As shown, the k-th grid-connected inverter includes a power conversion bridge, a PWM modulation stage, a GFMI controller, a filter impedance, and a virtual impedance. The GFMI controller generates a voltage modulation quantity, which is then applied to the power conversion bridge via PWM. The output voltage of the grid-connected inverter is connected to the AC network on the low-voltage side via the filter impedance and the virtual impedance (i.e., the virtual impedance introduced in this embodiment of the invention), and then supplies power to the load via the MV / HV network.

[0031] Specifically, a basic synchronization controller is introduced to synchronize the voltage of each grid-connected inverter. This basic synchronization controller can be a schedulable virtual oscillation controller based on contraction theory. After synchronization, the i-th grid-connected inverter is equivalent to a synchronization voltage source connected to a local equivalent impedance Z. i Ignoring voltage drop on the network and considering only load voltage drop, the equivalent admittance on the grid side is given as Y. net Local branch admittance Y i And the overall director Y Σ Defined as: ; .

[0032] According to Kirchhoff's laws, the voltage V at the PCC (point of common coupling) and the current I of the i-th grid-connected inverter... i satisfy: ; ; in, and Let represent the magnitude and phase angle of the synchronous voltage vector of the m-th grid-connected inverter, respectively. If all grid-connected inverters satisfy the shrinkage condition, then their output voltage... If they converge to the same voltage trajectory, then their current ratios satisfy: .

[0033] The above formula shows that after each grid-connected inverter meets the shrinkage requirement and its output voltage converges to the same voltage trajectory, its current distribution is mainly determined by the local equivalent impedance of the branches of each grid-connected inverter. Therefore, this embodiment of the invention introduces virtual impedances into the paths of each grid-connected inverter. To proactively shape Z i This allows for the allocation of a specified current ratio.

[0034] To determine the accurate value of the virtual impedance introduced on the branch of each grid inverter, embodiments of the present invention distinguish between two cases: the complete network topology is unknown and the complete network topology is known.

[0035] In cases where the complete network topology is unknown, the baseline virtual resistance and inductance corresponding to a single current share can be determined first. Then, combining the baseline virtual resistance and inductance corresponding to the single current share with the target current allocation weights of each grid inverter, the initial virtual impedance corresponding to each grid inverter is calculated according to an allocation rule inversely proportional to the target current allocation weights. Here, the baseline virtual resistance and inductance corresponding to a single current share can be randomly selected within a preset value range.

[0036] In some embodiments, the initial virtual impedance corresponding to each grid inverter can be calculated based on the following formula: ; ; in, and These are the reference virtual resistance and reference virtual inductance corresponding to a single current share, respectively. Assign weights to the target current of the k-th grid-connected inverter. and These represent the initial virtual resistance and initial virtual inductance corresponding to the k-th grid-connected inverter, and the initial virtual impedance corresponding to the k-th grid-connected inverter, respectively. From the initial virtual resistance and initial virtual inductance constitute: ω0 is the rated angular frequency.

[0037] It can be seen that the target current is assigned a weight ρ k The larger the value, the smaller the corresponding virtual impedance, and the larger the local equivalent admittance of the branch. The inverter in this grid structure naturally bears a larger output current. The target current allocation weight ρ k The smaller the value, the larger the corresponding virtual impedance, and the output current is suppressed.

[0038] Given a known complete network topology, initial virtual impedance scaling values ​​can be set for each grid inverter. The virtual impedance scaling values ​​for each inverter are adjusted to minimize the deviation between the predicted current share and the target current share under the current virtual impedance scaling values. Based on the adjusted optimized virtual impedance scaling values, the initial virtual impedance of each inverter is calculated. The optimization objective can be expressed as: , here To determine the predicted current share of the k-th grid-connected inverter under the current virtual impedance scaling value, even if the largest relative current distribution deviation among all inverters is minimized, the adjustment of the virtual impedance scaling value corresponding to each grid-connected inverter is stopped, thus obtaining the optimized virtual impedance scaling value corresponding to each grid-connected inverter. Based on the adjusted optimized virtual impedance scaling value corresponding to each grid-connected inverter... , can be based on (in Calculate the initial virtual impedance corresponding to the grid inverter.

[0039] When determining the predicted current share of the k-th (k=1,…,N) grid inverter under the current virtual impedance scaling value, the node admittance equation and the equivalent branch current equation of each grid inverter can be solved simultaneously based on the node admittance matrix corresponding to the current network topology to obtain the equivalent branch current vector corresponding to the current network topology. Based on the equivalent branch current of each grid inverter contained in the equivalent branch current vector, the predicted current share of each grid inverter under the current virtual impedance scaling value can be calculated.

[0040] In some embodiments, the nodal admittance equation is: ; The equivalent branch current equation for the kth grid-connected inverter is: ; Where Y is the node admittance matrix corresponding to the current network topology, and v is the access node voltage vector corresponding to the current network topology. Let v be the voltage at the access node corresponding to the k-th grid-connected inverter. This represents the equivalent branch current vector corresponding to the current network topology. for The equivalent branch current corresponding to the k-th grid inverter is Let be the output voltage of the k-th grid-connected inverter; where, according to contraction, the output voltages of all inverters should converge to the same voltage trajectory, so They are considered to be approximately the same; , and Here, ω0 represents the filter resistor and filter inductor of the k-th grid-connected inverter, and ω0 is the rated angular frequency. , and These are the reference virtual resistance and reference virtual inductance corresponding to a single current share, respectively. This is the current virtual impedance scaling value corresponding to the kth grid-connected inverter; After obtaining the equivalent branch current vector corresponding to the current network topology, the predicted current share of the k-th grid inverter under the current virtual impedance scaling condition is calculated based on the following formula. : ; Where N represents the total number of grid-connected inverters.

[0041] In other embodiments, the virtual impedance scaling value corresponding to any grid-connected inverter can be adjusted using the following formula: ; in, and These represent the virtual impedance scaling values ​​before and after adjustment for the k-th grid-connected inverter, respectively; η is the relaxation factor, typically 0 < η ≤ 1; the SAT function is a saturation function, and its output is limited to its minimum value. and maximum value The range is used as a numerical protection range to avoid the virtual impedance being too small or too large. The target current share for the k-th grid-connected inverter. For the k-th grid inverter in The predicted current share under the given conditions. This iterative formula is essentially a multiplicative fixed-point iterative formula with saturation constraints.

[0042] After obtaining the initial virtual impedance of each grid inverter using the above method, in order to further improve the accuracy of current ratio allocation, simulation or testing can be performed based on the initial virtual impedance of each grid inverter to obtain the actual current share of each grid inverter under the corresponding initial virtual impedance conditions. Then, based on the actual current share and target current share of each grid inverter, the initial virtual impedance of the corresponding grid inverter can be optimized to obtain the calibrated virtual impedance of each grid inverter that can achieve its target current share.

[0043] In some embodiments, after obtaining the initial virtual impedance when the complete network topology is unknown or when the complete network topology is known, the initial virtual impedance corresponding to each network inverter can be optimized based on the following formula: ; ; in, and These are the calibration virtual resistance and calibration virtual inductance corresponding to the k-th grid-connected inverter, and the calibration virtual impedance corresponding to the k-th grid-connected inverter, respectively. It consists of a calibrated virtual resistor and a calibrated virtual inductor. ), where γ is the correction index, and The initial virtual resistance and initial virtual inductance are obtained by decomposing the initial virtual impedance corresponding to the k-th grid-connected inverter. This represents the actual current share corresponding to the k-th grid-connected inverter. This represents the target current share of the k-th grid-connected inverter.

[0044] Using the above method, if the actual output of the k-th grid-connected inverter exceeds the target current share, then If the optimized virtual impedance increases, the branch current of the grid inverter will decrease; if the actual output of the k-th grid inverter is lower than the target current share, the optimized virtual impedance will decrease, thereby increasing the branch current of the grid inverter.

[0045] In some embodiments, when a phase-to-phase short circuit or asymmetrical voltage drop occurs between phases A and B, the current limiting input of each grid inverter is obtained. The current limiting input of any grid inverter... This is obtained by normalizing the maximum value of the three-phase output current of the grid-connected inverter. That is: ; in, , , Let K be the three-phase output current of the k-th grid-connected inverter. The rated current is used. The physical meaning of using the maximum phase output current is that: asymmetrical faults such as phase-to-phase short circuits and asymmetrical voltage drops between phases A and B will cause the current of one or two phases to exceed the limit first, and the maximum phase output current can directly trigger current limiting.

[0046] If the current limiting input of any grid inverter exceeds the preset current limiting target, the virtual impedance corresponding to the grid inverter will be increased by the PI controller based on its rated virtual impedance to limit the current limiting input of the grid inverter.

[0047] In some embodiments, the PI controller increases the virtual impedance corresponding to the grid inverter based on the following formula: ; ; ; in, The current limiting input for the k-th grid-connected inverter at time t. To preset the traffic limiting target, Let q be the current limiting error at time t. I,k (t) represents the state variable of the k-th grid-connected inverter corresponding to the integral element of the PI controller, and Δt is the controller sampling period. This is the virtual impedance corresponding to the k-th grid-connected inverter obtained after PI controller adjustment. The virtual impedance corresponding to the k-th grid-connected inverter is... and These are the proportional and integral coefficients of the PI controller, respectively. When Not exceeding hour, The virtual impedance does not increase additionally; when Exceed hour, It becomes the input to the PI control circuit.

[0048] Subsequently, the calibrated virtual impedance of the grid inverter or the virtual impedance corresponding to the grid inverter obtained after adjustment by the PI controller is incorporated into the voltage command of the grid inverter, so that the voltage command of the grid inverter satisfies: ; in, Let be the reference voltage that the k-th grid-connected inverter actually outputs to the voltage control loop at time t. Let be the synchronization voltage indicated by the synchronization voltage command at time t. This refers to the calibrated virtual impedance corresponding to the k-th grid-connected inverter, or the virtual impedance corresponding to the k-th grid-connected inverter after PI controller adjustment. Let be the output current vector of the k-th grid-connected inverter in the stationary αβ coordinate system. This is the virtual impedance drop calculated by the controller of the k-th grid inverter based on its output current vector and its calibrated virtual impedance or the virtual impedance adjusted by the PI controller. The physical meaning of this formula is as follows: the basic synchronous controller provides the synchronous voltage trajectory; the virtual impedance forms an adjustable virtual voltage drop at the voltage command layer; the optimized calibrated virtual impedance maintains the current ratio of each grid inverter under normal operating conditions; and the PI controller directly adjusts the virtual impedance during faults. Simultaneously, the controller of the k-th grid inverter sends this virtual impedance drop to its corresponding input to correct the internal voltage state change rate, ensuring that the dynamic equation of the output voltage of the k-th grid inverter satisfies the contracted Jacobian matrix without virtual impedance, thus achieving current limiting.

[0049] Specifically, to ensure that the contraction vector field does not contain the calibrated virtual impedance or the virtual impedance obtained from the PI controller, virtual impedance matching feed-in is required (that is, the virtual impedance voltage drop caused by the calibrated virtual impedance or the virtual impedance obtained from the PI controller is fed back to the corresponding controller to adjust the internal voltage state change rate of the k-th grid inverter). The reason is that if the virtual impedance voltage drop is only introduced into the voltage command of the grid inverter without matching feed-in, the Jacobian matrix of the output voltage dynamic equation of the grid inverter will increase the closed-loop voltage dynamics, including additional virtual impedance or the continuous PI integral state entering the voltage dynamics through the time-varying virtual impedance, which may cause the original contraction property to be lost due to the unmatched virtual impedance term.

[0050] Taking a shrinkable schedulable virtual oscillation controller with matching feedforward as an example, in the stationary αβ coordinate system, the controller generates the internal voltage state of the k-th grid inverter according to the following dynamic equation: ; ; ; in, and These are the α-axis and β-axis components of the internal voltage state of the k-th grid-connected inverter, respectively. For state-dependent amplitude recovery term, To recover the gain of the oscillator amplitude, This is the effective value of the rated voltage. The rated angular frequency, For current feedback gain, and It is an internal voltage state correction vector of the controller formed using output current feedback information, representing respectively α-axis components and β-axis components; The controller internal voltage state correction vector, generated based on the output current vector, is constructed according to the following formula: ; ; ; in, Let be the output current vector of the k-th grid-connected inverter in the stationary αβ coordinate system. and These are the filter inductor and the filter resistor, respectively. To calibrate the virtual impedance or the virtual impedance obtained from the PI controller, and They are respectively The virtual resistance and virtual inductance obtained from the decomposition; It is a second-order identity matrix. It is a second-order orthogonal rotation matrix; To reduce the virtual impedance voltage drop fed into the schedulable virtual oscillation controller. Because the voltage command of the grid inverter satisfies: Therefore, neglecting network voltage drop, the internal voltage state dynamics of the k-th grid-connected inverter satisfies: ; in Let be the voltage vector at the common coupling point of the k-th grid-connected inverter (ignoring network voltage drop). Substituting the controller's internal voltage state correction vector, generated from the output current vector, into the internal voltage state dynamic equation, we obtain the specific form of the internal voltage state dynamic equation for the k-th grid-connected inverter, ignoring network voltage drop: ; Calculating the Jacobian matrix of the right-hand function with respect to the internal voltage state reveals that the virtual resistance and virtual inductance do not appear in the aforementioned Jacobian matrix. Therefore, the contraction property is not affected by the virtual impedance.

[0051] In summary, the controller calculates the virtual impedance drop based on the calibrated virtual impedance of the k-th grid inverter or the virtual impedance adjusted by the PI controller and the corresponding output current vector of the grid inverter. The voltage command for the k-th grid-connected inverter is generated using this virtual impedance voltage drop. Simultaneously, the controller sends the virtual impedance voltage drop to the corresponding input of the controller's internal voltage state correction vector used to generate the output current, allowing the internal voltage state correction vector to participate in the control of the internal voltage state rate of change. Therefore, after substituting the controller's internal voltage state correction vector generated based on the output current vector into the internal voltage state dynamic equation, the virtual impedance voltage drop is eliminated from the internal voltage state dynamic equation. The resulting internal voltage state dynamic equation only includes the internal voltage state, the common coupling point voltage, and the virtual oscillation controller control parameters, no longer including the virtual resistance, virtual inductance, and the PI controller state that adjusts the virtual impedance. Therefore, the virtual impedance can play a role in current distribution and fault current limiting by correcting the actual voltage command, while not entering the Jacobian matrix related to the internal voltage state in the internal voltage state dynamic equation, thus preserving the virtual impedance's inherent contraction properties.

[0052] In other embodiments, the current limiting input of each grid-connected inverter is acquired in real time; If the current limiting input of any grid inverter is lower than the release threshold within the preset release time, the virtual impedance corresponding to that grid inverter will be reset to its corresponding calibrated virtual impedance, so that the system returns to the proportional distribution state.

[0053] In other embodiments, the current limiting target can be set to 1.5 pu, the release threshold can be set to 1.2 pu, and the preset release time can be set to 5 ms.

[0054] To verify the method proposed in the embodiments of the present invention, the following verification designs were performed: I. Normal operating condition allocation of three virtual impedance selection methods under random set A This embodiment presents a normal operating condition design where all 33 grid-connected inverters participate in the target current share allocation. The engineering connection structure is as follows: Figure 1 As shown, a random set A = {3, 9, 10, 19, 20, 21, 27, 28, 30, 31} is selected. For the 10 grid-connected inverters within set A, a target current allocation weight ρk = 2 is assigned, and for the remaining 23 inverters, a target current allocation weight ρk = 1 is assigned. The target current share is determined according to the proportion of each target current allocation weight to the total weight of the entire site. This setting indicates that the grid-connected inverters within set A bear twice the target current share, while the remaining grid-connected inverters bear one target current share. Figure 1 The virtual impedance Zvr,k(t) is located between the output of the grid inverter and the AC network, indicating that the configuration of the target current share in this embodiment ultimately affects the output current distribution through the difference in the virtual impedance of each grid inverter branch.

[0055] For this random set A, three virtual impedance selection paths can be adopted: First, when the complete network topology is unknown, use the baseline virtual impedance Z corresponding to a single current share. vr,0 Based on ρ k The first approach involves three steps: 1) configuring virtual impedance inversely, i.e., black-box inverse design in the first round; 2) when the complete network topology is known, using the node admittance equations and the equivalent branch current equations of each inverter in the network, along with the maximum distribution error target, as an alternative path to the first approach, generating virtual impedance based on the network model, i.e., node equation substitution design; 3) when model deviations or field deviations exist, using complete model simulation or field testing calibration, correcting the current share of each inverter in the network based on the selected virtual impedance. These three paths are related to... Figure 3 The three types of bar charts correspond one-to-one: green bars represent the first round of inverse proportional design of the black box, red bars represent the nodal equation substitution design, and blue bars represent the calibration results of the complete model. Figure 3 The left side shows full load PF=1.00, and the right side shows half load PF=1.00. The vertical values ​​represent the current distribution error of each grid-connected inverter relative to the target current share.

[0056] In full-load and half-load tests with a power factor PF=1.00, the maximum current distribution error corresponding to the first round of inverse proportional design in the black box was approximately 3.37% and 6.46%, respectively; the maximum current distribution error corresponding to the nodal equation substitution design was approximately 31.03% and 35.59%, respectively; and the maximum current distribution error corresponding to the complete model correction was approximately 0.85% and 4.27%, respectively. Therefore, Figure 2 This can illustrate the impact of the three virtual impedance selection paths on current distribution error in this embodiment, and show that the selection of virtual impedance can be determined from the three paths of black box first-round inverse proportional design, nodal equation substitution design, and complete model correction, depending on the availability of engineering information.

[0057] II. Current limiting due to voltage dips and ABC grounding faults The fault input in this embodiment corresponds to Figure 3 The voltage drop during the ABC ground fault shown in (b) is as follows: Figure 3 The fault range shown in (b) represents a three-phase symmetrical voltage drop. The load step and fault clearing moments are used to form the current-limiting recovery verification process. When an ABC ground fault occurs on the PCC side, the three-phase voltage of the PCC drops simultaneously, and the output current of the grid inverter may rise rapidly. In this embodiment, it is used as the current-limiting input. When it exceeds 1.5 pu, the current-limiting error is positive, and the PI loop directly adjusts the virtual impedance, thereby increasing the virtual voltage drop in the current feedback and suppressing the fault current.

[0058] The rate limiting and synchronization recovery results in this embodiment correspond to Figure 4 (a) Displays the maximum phase current limiting input Ipu,k and the 1.5pu current limiting threshold of all 33 grid-connected inverters; (b) Displays the virtual impedance change after PI regulation; (c) Displays the recovery of output voltage amplitude; (d) Displays frequency recovery. After fault clearing, if Ipu,k is lower than 1.2pu and remains below it for 5ms, the PI regulation loop will release the Zvr,k(t) back to its normal operating condition calibration value according to the release logic. The reason for using hysteresis and release time is to avoid frequent switching of Zvr,k(t) near the threshold. Figure 4 The voltage amplitude and frequency recovery curves are used to illustrate that the three-phase fault current limiting and the synchronization remain after recovery.

[0059] III. Current Limiting for AB Unbalanced Short Circuit Faults The fault input in this embodiment corresponds to Figure 3 The unbalanced short-circuit voltage drop shown in (a) is due to the AB phase short circuit being an asymmetrical fault; the currents in each phase may not reach their peak values ​​simultaneously. Figure 3 (a) is used to illustrate the different degrees of voltage drop in the three phases under unbalanced faults. In this embodiment, the maximum phase current is used to construct Ipu,k, so that current limiting can be directly triggered when any phase experiences overcurrent.

[0060] In the asymmetric fault current limiting mode, the PI circuit directly adjusts the virtual impedance Zvr,k(t) of each grid-connected inverter and performs local current limiting via the voltage command. By using the same current limiting target, saturation upper limit, and slope constraint to limit the virtual impedance variation of each inverter, the virtual impedance feedforward term acts on the voltage command in a constrained manner. The control effect of this embodiment corresponds to... Figure 5 (a) Displays the maximum phase current limiting process, (b) Displays the virtual impedance adjustment process, (c) Displays voltage amplitude recovery, and (d) Displays frequency recovery. The basic synchronous controller remains responsible for maintaining the output voltage trajectory convergence, while the virtual impedance is responsible for generating the current-limiting virtual voltage drop. Figure 5 This demonstrates that under unbalanced faults, "current limiting and synchronization" can be achieved by directly adjusting the virtual impedance using a PI converter.

[0061] In summary, the method provided by this invention, when the complete network topology is unknown, calculates the initial virtual impedance of each grid inverter based on the reference virtual resistance and reference virtual inductance corresponding to a single current share and the target current allocation weight of each grid inverter. An implementable first-round current proportional allocation can be obtained by inversely proportionalizing the virtual impedance with the target current allocation weight. When the complete network topology is known, the initial virtual impedance scaling value corresponding to each grid inverter is set, and the virtual impedance scaling value corresponding to each grid inverter is adjusted with the goal of minimizing the deviation between the predicted current share and the target current share of each grid inverter under the current virtual impedance scaling value. Based on the adjusted optimized virtual impedance scaling value, the initial virtual impedance of the corresponding grid inverter is calculated as an alternative path to black-box parameter tuning. Furthermore, through simulation or field testing calibration, the initial virtual impedance of the corresponding grid inverter is optimized based on the actual current share and target current share of each grid inverter. This reduces the residual allocation error in the complete site model and improves the accuracy of proportional current allocation in large-scale grid inverter scenarios.

[0062] Furthermore, during faults such as PCC voltage drop, ABC grounding fault, and AB unbalanced short circuit, by acquiring the current limiting input of each grid inverter, when the current limiting input of any grid inverter exceeds the preset current limiting target, the virtual impedance corresponding to that grid inverter is increased using a PI controller based on its calibrated virtual impedance to limit the current limiting input of that grid inverter. This can quickly increase the equivalent impedance, limit overcurrent, and avoid sudden surges after fault clearance. Simultaneously, the grid inverter's controller sends the virtual impedance voltage drop corresponding to the calibrated virtual impedance or the virtual impedance adjusted by the PI controller to the corresponding input terminal of the controller to correct the internal voltage state change rate. This ensures that the contracted Jacobian matrix in the dynamic equation of the output voltage of the corresponding grid inverter does not contain virtual impedance, thereby enabling the grid inverter to achieve current proportional distribution under normal operating conditions and maintain output voltage synchronization during fault current limiting.

[0063] This invention provides a current proportional distribution and current limiting controller for multiple grid-connected inverters, which is used to execute the methods described in any of the above embodiments.

[0064] This invention provides an electronic device, including: a computer-readable storage medium and a processor; The computer-readable storage medium is used to store executable instructions; The processor is configured to read executable instructions stored in the computer-readable storage medium and execute the method as described in any of the above embodiments.

[0065] This invention provides a computer-readable storage medium storing computer instructions that cause a processor to perform the method described in any of the above embodiments.

[0066] This invention provides a computer program product, including a computer program or instructions, which, when executed by a processor, implement the method described in any of the above embodiments.

[0067] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit 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 proportional current distribution and current limiting control of multiple grid-connected inverters, characterized in that, include: S1, determine the target current allocation weight and target current share of each grid inverter; S2, when the complete network topology is unknown, calculate the initial virtual impedance of each network inverter based on the reference virtual resistance and reference virtual inductance corresponding to the single current share and the target current allocation weight of each network inverter. When the complete network topology is known, the initial virtual impedance scaling value corresponding to each grid inverter is set, and the virtual impedance scaling value corresponding to each grid inverter is adjusted with the goal of minimizing the deviation between the predicted current share and the target current share of each grid inverter under the current virtual impedance scaling value. Based on the optimized virtual impedance scaling value corresponding to each grid inverter obtained by adjustment, the initial virtual impedance corresponding to the corresponding grid inverter is calculated. The predicted current share of any grid-connected inverter under the current virtual impedance scaling condition is determined based on the following method: Based on the node admittance matrix corresponding to the current network topology, the node admittance equation and the equivalent branch current equation of each grid inverter are solved simultaneously to obtain the equivalent branch current vector corresponding to the current network topology. Based on the equivalent branch current of each grid inverter contained in the equivalent branch current vector, the predicted current share of each grid inverter under the current virtual impedance scaling value is calculated. S3. Based on the initial virtual impedance corresponding to each grid inverter, perform a complete site model simulation or test to obtain the actual current share corresponding to each grid inverter. Based on the actual current share and target current share corresponding to each grid inverter, optimize the initial virtual impedance corresponding to the corresponding grid inverter to obtain the calibrated virtual impedance corresponding to each grid inverter that can achieve its target current share. S4, the controller of the k-th grid inverter obtains the virtual impedance drop based on the output current vector and its calibrated virtual impedance; the controller stores the virtual impedance drop, uses it to generate the voltage command of the k-th grid inverter, and sends the virtual impedance drop to the corresponding input terminal of the controller to correct the internal voltage state change rate.

2. The current proportional distribution and current limiting control method for multiple grid-connected inverters as described in claim 1, characterized in that, The nodal admittance equation is: ； The equivalent branch current equation for the kth grid-connected inverter is: ； Where Y is the node admittance matrix corresponding to the current network topology, and v is the access node voltage vector corresponding to the current network topology. Let v be the voltage at the access node corresponding to the k-th grid-connected inverter. This represents the equivalent branch current vector corresponding to the current network topology. for The equivalent branch current corresponding to the k-th grid inverter is The output voltage of the k-th grid-connected inverter; , and Here, ω0 represents the filter resistor and filter inductor of the k-th grid-connected inverter, and ω0 is the rated angular frequency. , and These are the reference virtual resistance and reference virtual inductance corresponding to a single current share, respectively. This is the current virtual impedance scaling value corresponding to the kth grid-connected inverter; After obtaining the equivalent branch current vector corresponding to the current network topology, the predicted current share of the k-th grid inverter under the current virtual impedance scaling condition is calculated based on the following formula. : ； Where N represents the total number of grid-connected inverters.

3. The method for proportional current distribution and current limiting control of multiple grid-connected inverters as described in claim 1, characterized in that, The virtual impedance scaling value corresponding to any grid-connected inverter is adjusted according to the following formula: ； in, and These represent the virtual impedance scaling values ​​before and after adjustment for the k-th grid-connected inverter, respectively; η is the relaxation factor, and the sat function is a saturation function whose output is limited to its minimum value. and maximum value between, The target current share for the k-th grid-connected inverter. For the k-th grid inverter in Predicted current share under certain conditions.

4. The method for proportional current distribution and current limiting control of multiple grid-connected inverters as described in claim 1, characterized in that, When the complete network topology is unknown, the initial virtual impedance of each network inverter is calculated based on the following formula: ； ； in, and These are the reference virtual resistance and reference virtual inductance corresponding to a single current share, respectively. Assign weights to the target current of the k-th grid-connected inverter. and These are the initial virtual resistance and initial virtual inductance corresponding to the k-th grid inverter, respectively. The initial virtual impedance corresponding to the k-th grid inverter is composed of the initial virtual resistance and initial virtual inductance.

5. The method for proportional current distribution and current limiting control of multiple grid-connected inverters as described in claim 1, characterized in that, After obtaining the initial virtual impedance when the complete network topology is unknown or known, the initial virtual impedance corresponding to each network inverter is optimized based on the following formula: ； ； in, and These are the calibration virtual resistor and calibration virtual inductor corresponding to the k-th grid-connected inverter, respectively. The calibration virtual impedance corresponding to the k-th grid-connected inverter is composed of the calibration virtual resistor and calibration virtual inductor. γ is the correction index. and The initial virtual resistance and initial virtual inductance are obtained by decomposing the initial virtual impedance corresponding to the k-th grid-connected inverter. This represents the actual current share corresponding to the k-th grid-connected inverter. This represents the target current share of the k-th grid-connected inverter.

6. The method for proportional current distribution and current limiting control of multiple grid-connected inverters as described in any one of claims 1 to 5, characterized in that, The method further includes: When a phase-to-phase short circuit or asymmetrical voltage drop occurs between phases A and B, the current limiting input of each grid inverter is obtained; wherein, the current limiting input of any grid inverter is obtained by normalizing the maximum value of the three-phase output current of the grid inverter. If the current limiting input of any grid inverter exceeds the preset current limiting target, the virtual impedance of the grid inverter will be increased by the PI controller based on its rated virtual impedance to limit the current limiting input of the grid inverter.

7. The method for proportional current distribution and current limiting control of multiple grid-connected inverters as described in claim 6, characterized in that, The PI controller increases the virtual impedance of the grid-connected inverter based on the following formula: ； ； ； in, The current limiting input for the k-th grid-connected inverter at time t. For the preset rate limiting target, Let q be the current limiting error at time t. I,k (t) represents the state variable of the k-th grid-connected inverter corresponding to the integral element of the PI controller, and Δt is the controller sampling period. The virtual impedance corresponding to the k-th grid-connected inverter obtained after adjustment by the PI controller. The virtual impedance corresponding to the k-th grid-connected inverter is... and These are the proportional and integral coefficients of the PI controller, respectively.

8. The method for proportional current distribution and current limiting control of multiple grid-connected inverters as described in claim 6, characterized in that, The method of increasing the virtual impedance corresponding to the grid inverter through a PI controller to limit the current limiting input of the grid inverter further includes: Real-time acquisition of current limiting input from each grid-connected inverter; If the current limiting input of any grid inverter is lower than the release threshold within the preset release time, the virtual impedance corresponding to that grid inverter will be reset to its corresponding calibrated virtual impedance.

9. The method for proportional current distribution and current limiting control of multiple grid-connected inverters as described in claim 6, characterized in that, The method further includes: The virtual impedance corresponding to each grid inverter is incorporated into the voltage command of that grid inverter, so that the voltage command of that grid inverter satisfies: ； in, Let be the reference voltage that the k-th grid-connected inverter actually outputs to the voltage control loop at time t. Let be the synchronization voltage indicated by the synchronization voltage command at time t. This refers to either the calibrated virtual impedance corresponding to the k-th grid-connected inverter or the virtual impedance corresponding to the k-th grid-connected inverter after adjustment by the PI controller. Let be the output current vector of the k-th grid-connected inverter in the stationary αβ coordinate system.

10. A current proportional distribution and current limiting controller for multiple grid-connected inverters, characterized in that, The controller is used to execute the current proportional distribution and current limiting control method for multiple grid-connected inverters as described in any one of claims 1 to 9.

Images