A photovoltaic direct-current power generation system collaborative optimization control method based on virtual damping injection

Abstract

The present application relates to the technical field of direct current power transmission grid and equipment, and discloses a photovoltaic direct current power generation system collaborative optimization control method based on virtual damping injection, which comprises the following steps: determining an oscillation risk, outputting reverse harmonic current by a daytime group in cooperation with a nighttime group, and sending an oscillation suppression request if the total margin of energy storage current of the daytime group and the nighttime group is still insufficient; detecting low-voltage energy storage current margin of interconnected branches, and performing collaborative harmonic suppression when the margin is satisfied; when the margin of all interconnected branches is not satisfied, calculating the required reduced active power, reducing the power of the daytime group energy storage, and then limiting the power of the remaining part according to the harmonic voltage content sorting of the photovoltaic power generation unit; starting local-level optimization, charging the energy storage by the daytime group in cooperation with the nighttime group, and starting collaborative-level optimization control strategy when the energy storage cannot be charged; and screening target branches and performing cross-branch power transfer. The present application significantly improves the operation stability and self-adaptive ability of the photovoltaic direct current power generation unit sending-out system under complex working conditions.

Description

Technical Field

[0001] This invention relates to the field of DC power transmission grid and equipment technology, and more specifically, to a collaborative optimization control method for photovoltaic DC power generation systems based on virtual damping injection. Background Technology

[0002] New energy aggregation can employ two technical solutions: AC aggregation and DC aggregation. AC aggregation, with its mature technology, is currently the mainstream solution in engineering projects. However, in desert and barren areas located at the end of the AC power grid, the system strength is insufficient, and AC aggregation faces issues such as power angle stability. In contrast, DC aggregation does not require consideration of grid strength and possesses active voltage support capabilities, making it particularly suitable for scenarios with a wide aggregation range and long transmission distance at the end of the AC power grid.

[0003] To provide sufficient capacity and power utilization for a single DC transformer, photovoltaic (PV) power generation units are connected in parallel on the low-voltage side of the DC transformer, and their capacity is larger than that of the DC transformer. When sunlight is intense, the PV power generation units need to perform power transfer or power limiting. Therefore, it is necessary to study local-level and system-level coordinated control technologies for the PV power generation units. A virtual impedance injection method is proposed to achieve coordinated control. Simultaneously, the PV power generation units exhibit different negative damping states under different operating modes to prevent oscillations on the low-voltage DC bus. The DC transformer needs to detect oscillation characteristics and initiate virtual damping control to suppress oscillations. Therefore, to ensure the safe and stable operation of the grid-connected system, a coordinated optimization control method for PV DC power generation systems based on virtual damping injection is urgently needed.

[0004] No effective solutions have yet been proposed to address the problems in the relevant technologies. Summary of the Invention

[0005] To address the problems in related technologies, this invention proposes a collaborative optimization control method for photovoltaic DC power generation systems based on virtual damping injection, in order to overcome the aforementioned technical problems existing in the existing related technologies.

[0006] Therefore, the specific technical solution adopted by the present invention is as follows:

[0007] A collaborative optimization control method for a photovoltaic DC power generation system based on virtual damping injection, the method comprising:

[0008] S1. Extract the DC voltage of the low-voltage DC bus from the DC transformer and determine the oscillation risk. If there is a risk, calculate the required reverse harmonic current and output the reverse harmonic current by the day group and the night group. If the total margin of the energy storage current of the day group and the night group is still insufficient, send an oscillation suppression request to the DC collaborative control system.

[0009] S2. After receiving the oscillation suppression request, the DC collaborative control system detects the low-voltage energy storage current margin of the interconnected branches. If the margin is met, collaborative harmonic suppression is performed. If the margin of all interconnected branches is not met, the required reduction of active power is calculated. After the daytime energy storage reduces the power, the remaining part is limited by the photovoltaic power generation units according to the harmonic voltage content.

[0010] S3, in parallel with the S2 oscillation suppression process, the DC transformer detects the low-voltage side input current during periods of sufficient sunlight. When the current exceeds the limit, it sends a power limit request to the DC collaborative control system and initiates local-level optimization. The daytime group and the nighttime group work together to charge the energy storage. If charging is not possible, the collaborative-level optimization control strategy is initiated.

[0011] S4. After the collaborative optimization control strategy is started, the harmonic content and active damping injection status of each interconnected low-voltage bus are detected, the target branch is screened and cross-branch power transfer is performed; if the target branch can only partially take over, the remaining power is absorbed by the low-voltage energy storage of the interconnected branch. If there is still no effective transfer path, the power is reduced by virtual impedance control. After completion, it enters normal monitoring.

[0012] S5. When a DC transformer fault is detected during routine monitoring, the low-voltage energy storage is disconnected from the grid to maintain the bus voltage. After the bus stabilizes, power transfer coordination is performed. After the faulty transformer recovers, the power transfer command is withdrawn, and each power generation unit exits the virtual damping distribution and resumes natural power transmission.

[0013] The beneficial effects of this invention are as follows:

[0014] 1. When the output power of the DC transformer is limited, the present invention uses the diode disconnect switch of the low-voltage bus of the DC transformer to interconnect multiple DC power generation units. Based on the principle of prioritizing local collaborative control and optimizing system collaborative control, the surplus power is transferred by prioritizing local low-voltage energy storage charging. The power is automatically distributed among multiple DC power generation units by actively injecting virtual resistors, thereby realizing the transfer of surplus power.

[0015] 2. This invention enables automatic distribution of surplus power among DC power generation units interconnected by diode isolating switches by activating local and system-level system control when power is limited and actively injecting virtual damping to change the characteristics of the DC transformer. When the photovoltaic power generation voltage exhibits different negative damping characteristics in multiple operating modes, in order to avoid oscillation, the AC component of the low-voltage bus is checked and actively injected virtual impedance is activated to suppress oscillation, thereby significantly improving the operational stability and adaptability of the photovoltaic DC power generation unit transmission system under complex operating conditions. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a schematic diagram of a photovoltaic DC power generation system in a collaborative optimization control method for a photovoltaic DC power generation system based on virtual damping injection, according to an embodiment of the present invention.

[0018] Figure 2 This is a schematic diagram of power transfer at the local level and system level in a collaborative optimization control method for a photovoltaic DC power generation system based on virtual damping injection, according to an embodiment of the present invention.

[0019] Figure 3 This is a schematic diagram of the power-limited virtual impedance injection process in a collaborative optimization control method for a photovoltaic DC power generation system based on virtual damping injection according to an embodiment of the present invention.

[0020] Figure 4 This is a schematic flowchart of a low-voltage bus oscillation suppression method for a photovoltaic DC power generation system based on virtual damping injection, according to an embodiment of the present invention.

[0021] Figure 5 This is a schematic diagram of virtual impedance control in a collaborative optimization control method for a photovoltaic DC power generation system based on virtual damping injection, according to an embodiment of the present invention.

[0022] Figure 6 This is a flowchart of a collaborative optimization control method for a photovoltaic DC power generation system based on virtual damping injection, according to an embodiment of the present invention. Detailed Implementation

[0023] To further illustrate the various embodiments, the present invention provides accompanying drawings, which are part of the disclosure of the present invention. These drawings are mainly used to illustrate the embodiments and can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. With reference to these drawings, those skilled in the art should be able to understand other possible implementation methods and the advantages of the present invention.

[0024] According to an embodiment of the present invention, a collaborative optimization control method for a photovoltaic DC power generation system based on virtual damping injection is provided.

[0025] The present invention will now be further described in conjunction with the accompanying drawings and specific embodiments, such as... Figure 1 and Figure 6 As shown, according to an embodiment of the present invention, a collaborative optimization control method for a photovoltaic DC power generation system based on virtual damping injection includes:

[0026] S1. Extract the DC voltage of the low-voltage DC bus from the DC transformer and determine the oscillation risk. If there is a risk, calculate the required reverse harmonic current and output the reverse harmonic current by the day group and the night group. If the total energy storage current margin of the day group and the night group is still insufficient, send an oscillation suppression request to the DC collaborative control system.

[0027] The photovoltaic DC power generation system consists of a unidirectional DC transformer, photovoltaic DC power generation units (i.e., photovoltaic power generation units), an AC / DC converter, low-voltage energy storage, and diode disconnect switches. Multiple photovoltaic DC power generation units and low-voltage energy storage are connected in parallel to the low-voltage DC bus of the unidirectional DC transformer, then connected to the high-voltage DC bus via the DC transformer, and finally connected to the AC grid via AC / DC. The low-voltage DC bus of the DC transformer is flexibly interconnected via bidirectional diodes. To improve the utilization rate of the DC transformer, the total capacity of the photovoltaic power generation units is greater than the capacity of the DC transformer. The low-voltage energy storage is divided into daytime and nighttime groups. The nighttime group's low-voltage energy storage is charged to its maximum SOC during the day to increase the power generation time of the DC power generation system. The daytime group aims to mitigate the output of new energy sources. The flexible interconnection of bidirectional diodes utilizes the unidirectional conductivity of diodes to meet the premise of mutual assistance and prevent the spread of low-voltage bus faults to interconnected low-voltage buses.

[0028] S1 includes:

[0029] S11. The DC voltage of the low-voltage DC bus is collected in real time through the DC transformer, and the AC voltage component is extracted through the high-pass filter. Based on the AC voltage component, it is determined whether there is a risk of DC oscillation in the low-voltage DC bus. If so, the required reverse harmonic current is calculated.

[0030] S12. Check whether the daytime group's energy storage current margin capacity meets the required reverse harmonic current. If it does, allocate based on the key indicators of energy storage module temperature and equivalent internal resistance, with maintaining optimal temperature as the first priority and minimizing equivalent damping as the second priority, and allocate the output harmonic voltage, so that the daytime group outputs the reverse harmonic current.

[0031] S13. If the energy storage current margin of the daytime group cannot meet the demand, the energy storage current margin of the nighttime group will be further checked. If the total energy storage current margin of the nighttime group and the daytime group meets the demand, the harmonic output command will be allocated according to the same priority rule as S23, and the nighttime group and the daytime group will output the reverse harmonic current in concert.

[0032] S14. If the total energy storage current margin of the night group and the day group is still insufficient to meet the demand, an oscillation suppression request is sent to the DC collaborative control system to trigger the collaborative suppression of DC bus oscillation risk strategy.

[0033] It should be noted that, as Figure 4As shown, the photovoltaic power generation unit and low-voltage energy storage are connected in parallel on the low-voltage DC bus. The photovoltaic power generation unit and low-voltage energy storage are power sources and exhibit different negative damping characteristics. The DC transformer detects the DC voltage of the low-voltage DC bus in real time and extracts the AC voltage component through a high-pass filter. When the detected AC component is greater than the set threshold (preferably 5%) and shows an amplifying trend (i.e., the AC voltage component detected in this control cycle is greater than that in the previous control cycle), it is determined that there is a risk of DC oscillation on the low-voltage DC bus, and the required output reverse current is calculated.

[0034] ;

[0035] In the formula, R is the reverse harmonic current; kd is the harmonic damping gain coefficient; V +L V For the required virtual impedance; V thd For the extracted harmonics.

[0036] When the DC transformer detects that the low-voltage energy storage current margin of the daytime group is greater than the required reverse output current in Equation 1, the daytime group energy storage suppresses DC oscillations and allocates the output harmonic voltage based on key indicators such as energy storage module temperature and equivalent internal resistance. Maintaining optimal temperature (around 25℃) is the first priority, and low equivalent damping is the second priority. When the energy storage current margin of the daytime group is less than the required reverse harmonic current in Equation 1, the DC transformer further determines the energy storage current margin capacity of the nighttime group. If the nighttime group and the daytime group meet the required reverse harmonic current in Equation 1, the energy storage of the daytime group and the nighttime group work together to suppress the harmonic current. Based on key indicators such as energy storage module temperature and equivalent internal resistance, the output harmonic voltage is allocated based on key indicators such as energy storage module temperature and equivalent internal resistance. Maintaining optimal temperature (around 25℃) is the first priority, and low equivalent damping is the second priority.

[0037] Among them, when the illumination is uneven, the photovoltaic power generation unit with sufficient illumination is in the maximum power tracking mode, while the photovoltaic power generation unit with weak illumination is based on the PV curve and moves to the right half plane to seek the photovoltaic panel output voltage corresponding to the power limit value to adjust the boost duty cycle. The photovoltaic power generation unit with maximum power tracking, low-voltage energy storage and power-limited photovoltaic power generation unit are connected in parallel on the low-voltage DC bus and exhibit different negative damping characteristics in parallel.

[0038] A unidirectional DC transformer continuously monitors the DC voltage of the low-voltage DC bus and detects the AC voltage component of the DC bus using a high-pass filter. When the detected AC voltage component of the low-voltage DC bus exceeds a set threshold, it is determined that there is a risk of DC oscillation on the low-voltage DC bus, and damping control is initiated to suppress the oscillation, as shown in the following formula:

[0039] ;

[0040] In the formula, id is the current command to be superimposed; s is the Laplace operator; kd is the harmonic damping gain coefficient; h V is the high-pass cutoff angular frequency, and V is the AC component.

[0041] When photovoltaic power generation units are in the same operating mode, they exhibit consistent negative damping characteristics on the parallel low-voltage bus, and the DC transformer exits the oscillation suppression strategy.

[0042] S2. After receiving the oscillation suppression request, the DC collaborative control system detects the low-voltage energy storage current margin of the interconnected branches. If the margin is met, collaborative harmonic suppression is performed. If the margin of all interconnected branches is not met, the required reduction of active power is calculated. After the daytime energy storage reduces the power, the remaining part is limited by the photovoltaic power generation units according to the harmonic voltage content.

[0043] S2 includes:

[0044] S21. After receiving the oscillation suppression request, the DC cooperative control system detects the current margin of the low-voltage energy storage of the DC transformer branch interconnected by the diode isolating switch. If the current margin meets the required reverse harmonic current output, the forward transfer isolating switch will be closed to achieve cooperative harmonic suppression.

[0045] S22. If the current margin of all branch low-voltage energy storage cannot meet the required output reverse harmonic current, calculate the required reduction in active power and detect the active power output of daytime group energy storage.

[0046] S23. If the active power output of the daytime energy storage group meets the derating requirement, the allocation is based on the key indicators of energy storage module temperature and equivalent damping, with the highest temperature as the first priority and the maximum equivalent damping as the second priority, and the power reduction is carried out by the daytime group to reduce the active power.

[0047] S24. If the active power output of the daytime energy storage group is insufficient to cover all the derating requirements, the active power of the daytime group shall be reduced first, and then the remaining active power to be reduced shall be limited by the DC power generation units ranked first in terms of output active power. Among the DC power generation units participating in the power limiting, they shall be ranked according to the harmonic voltage content (harmonic components) of the photovoltaic output port, and the photovoltaic units ranked first in terms of harmonic factor shall be selected to reduce active power first.

[0048] S24 includes:

[0049] S241. If the active power output by the daytime group is less than the required reduction in active power, the daytime group shall reduce the active power first, and then the remaining active power deficit shall be calculated.

[0050] S242. Based on the remaining active power deficit, a probing power disturbance is applied sequentially to the daytime group and the photovoltaic power generation unit, and the change in system damping ratio is measured simultaneously. Based on the change in damping ratio, the unit power damping influence coefficient of the daytime group and the photovoltaic power generation unit is calculated.

[0051] S242 includes:

[0052] S2421. Determine the amplitude of the detection disturbance based on the remaining active power deficit, apply short-time pulse power disturbances to the daytime group and photovoltaic power generation unit in sequence, and simultaneously collect the voltage and current waveforms of the low-voltage DC bus.

[0053] S2422. Preprocess the voltage and current waveforms before and after the disturbance, and use a matrix beam strategy to extract the damping ratio of the dominant oscillation mode of the system. Record the damping ratio before the disturbance and the damping ratio after the disturbance respectively.

[0054] It should be noted that the mode identification matrix in this step is the Hankel matrix, which has the structural characteristic of equal elements along the antidiagonal and is constructed by arranging the sampling sequence of oscillating AC components with row-by-row delay. This invention employs a matrix bundle strategy based on the mode identification matrix to extract the damping ratio of the dominant oscillation mode of the system. This method directly performs pole decomposition on the measured voltage waveform without relying on the system's mathematical model or network parameters. By constructing the mode identification matrix and solving the matrix bundle eigenvalues, it can simultaneously and accurately extract the attenuation factor and oscillation angular frequency of each oscillation mode, showing significant advantages for damping identification under short data lengths. Simultaneously, the matrix bundle algorithm uses singular value decomposition to distinguish between the signal subspace and the noise subspace, effectively suppressing measurement noise interference. Furthermore, it automatically selects the dominant oscillation mode by solving the complex amplitude using the least squares method and selecting the mode with the largest complex amplitude as the criterion, avoiding the subjectivity of manual judgment. The calculation process requires no iteration, is highly efficient, and is suitable for online real-time execution.

[0055] S2422 includes:

[0056] S24221. Extract the low-voltage DC bus voltage sampling sequences for the corresponding durations before and after the disturbance, remove the DC bias component, and extract several oscillating AC component sequences as input signals.

[0057] It should be noted that S24221 includes: taking the time of disturbance application as the dividing point, extracting a period of voltage sampling data before the disturbance as the pre-disturbance sequence, and extracting a period of voltage sampling data after the disturbance as the post-disturbance sequence; the extraction period is 10 to 20 oscillation cycles of the expected dominant oscillation mode of the system; calculating the arithmetic mean of the pre-disturbance sequence and the post-disturbance sequence respectively, subtracting the average value from each sampling point of the original sequence to remove the DC bias component, and obtaining an AC component sequence containing only oscillation characteristics; and using the pre-disturbance AC component sequence and the post-disturbance AC component sequence as two independent input signals for subsequent mode identification.

[0058] S24222. Set the total number of sampling points and bundle parameters for each group of input signals, and construct mode identification matrices that differ by one sampling step for each oscillating AC component sequence.

[0059] It should be noted that S24222 includes:

[0060] Determine the total number of sampling points N for each group of input signals, which is the number of sampling points contained in the AC component sequence of that group;

[0061] Set the bundle parameter L, where L is an integer between N / 3 and N / 2. L determines the number of columns in the constructed matrix. The larger the number of columns, the higher the recognition accuracy and the greater the computational load.

[0062] Taking the unperturbed sequence as an example, it is arranged into a matrix with NL rows and L+1 columns: the first row takes the first to L+1 sampling points of the sequence, the second row takes the second to L+2 sampling points of the sequence, and so on, for a total of NL rows;

[0063] Take the first L columns of the above matrix as the preorder modality identification matrix, and take the last L columns as the postorder modality identification matrix; each row element of the postorder matrix is ​​shifted backward by one sampling step relative to the corresponding row of the preorder matrix, and the two form a matrix bundle.

[0064] Construct the corresponding preceding mode identification matrix and following mode identification matrix for the perturbed sequence in the same manner.

[0065] S24223. Solve for the pseudo-inverse matrix of the preceding mode identification matrix corresponding to each group of input signals, multiply the pseudo-inverse matrix with the following mode identification matrix, solve for all non-zero eigenvalues ​​of the product matrix, and obtain the pole sequence corresponding to each oscillation mode.

[0066] It should be noted that S24223 includes: performing singular value decomposition on the preceding mode identification matrix, constructing its pseudo-inverse matrix using the singular values ​​and left and right singular vectors; the dimension of the pseudo-inverse matrix is ​​the same as the dimension of the transposed preceding matrix; multiplying the pseudo-inverse matrix with the subsequent mode identification matrix to obtain a product matrix of dimension L×L; this product matrix contains the pole information of each mode in the oscillating signal; solving for all eigenvalues ​​of this product matrix, removing eigenvalues ​​that are zero or close to zero, and retaining non-zero eigenvalues; each non-zero eigenvalue corresponds to a pole of an oscillating mode, and all poles are arranged according to the contribution of the corresponding mode to form a pole sequence.

[0067] S24224. Calculate the attenuation factor and oscillation angular frequency of each oscillation mode under the corresponding working condition based on each set of pole sequences. Solve for the complex amplitude of each oscillation mode using the least squares method and select the oscillation mode with the largest complex amplitude as the dominant oscillation mode.

[0068] It should be noted that S24224 includes: For each pole z, the attenuation factor α and oscillation angular frequency ω are calculated as follows: α = ln|z| / Ts; ω = arctan[Im(z) / Re(z)] / Ts; where Ts is the sampling period, ln represents the natural logarithm, |z| is the pole magnitude, Im(z) and Re(z) are the imaginary and real parts of the pole, respectively, and arctan is the arctangent function; after calculating the attenuation factor and oscillation angular frequency of each oscillation mode, the original oscillating AC component sequence is fitted using the obtained pole sequence through the least squares method to solve for the complex amplitude corresponding to each mode. The least squares method determines the optimal estimate of each complex amplitude by minimizing the sum of squared errors between the fitted sequence and the actual sequence; the magnitudes of the complex amplitudes of each oscillation mode are compared, and the mode with the largest complex amplitude magnitude is selected as the dominant oscillation mode of the system under the current operating conditions.

[0069] S24225. Based on the attenuation factor and oscillation angular frequency of the dominant oscillation mode, calculate and record the damping ratio of the dominant oscillation mode of the system before and after the disturbance.

[0070] It should be noted that the attenuation factor and oscillation angular frequency are obtained from the pole information of the dominant oscillation mode. The attenuation factor and oscillation angular frequency are substituted into the damping ratio calculation relationship to obtain the damping ratio value. The above calculation is performed on the dominant oscillation mode obtained from the data before the disturbance to obtain the damping ratio before the disturbance and recorded. The above calculation is performed on the dominant oscillation mode obtained from the data after the disturbance to obtain the damping ratio after the disturbance and recorded.

[0071] S2423. Calculate the damping ratio change corresponding to the daytime group and photovoltaic power generation unit. The damping ratio change is the difference between the damping ratio after the disturbance and the damping ratio before the disturbance.

[0072] S2424. Based on the change in damping ratio and the corresponding applied disturbance power amplitude, calculate the unit power damping influence coefficient of the daytime group and the photovoltaic power generation unit, respectively.

[0073] It should be noted that the damping ratio change and disturbance power amplitude are extracted one by one from the daytime energy storage modules and photovoltaic power generation units that have completed the disturbance test. The damping ratio change obtained by the same unit in the same disturbance test is used as the numerator and the corresponding applied disturbance power amplitude is used as the denominator for the ratio calculation. The ratio result is used as the unit power damping influence coefficient of the unit under the current operating condition. The above calculation is repeated for each unit that has completed the disturbance test to obtain the unit power damping influence coefficient of all tested units and record it.

[0074] S243. Based on the unit power damping influence coefficient, the photovoltaic power generation unit is divided into a strong influence zone, a weak influence zone, and a negligible influence zone. Considering the energy storage state of charge and the zoning results, units in the weak influence zone and the negligible influence zone are prioritized to bear the power reduction, and the total power reduction of the photovoltaic power generation unit is determined, specifically including:

[0075] The first step is to take the unit power damping influence coefficient calculated for each photovoltaic power generation unit, and sort all the photovoltaic power generation units participating in the evaluation from largest to smallest according to the absolute value of their unit power damping influence coefficient to form a damping influence coefficient sequence.

[0076] The second step is to set the threshold values ​​for the first and second zones. The specific values ​​of the two thresholds are tuned offline based on the system's allowable damping reduction margin. Units with an absolute value of the unit power damping influence coefficient greater than the first threshold are classified into the strong influence zone, indicating that the power reduction of the unit has a significant negative impact on the system damping. Units with an absolute value between the first and second thresholds are classified into the weak influence zone, indicating that the power reduction of the unit has a certain impact on the system damping, but the degree is controllable. Units with an absolute value less than the second threshold are classified into the negligible zone, indicating that the power reduction of the unit has an approximately negligible impact on the system damping.

[0077] The third step is to read the current state of charge (SOC) value of the daytime energy storage module and determine whether it is higher than the preset safety lower limit. If the SOC is higher than the safety lower limit, it indicates that the daytime energy storage module has the capacity to further reduce power. At this time, the zoning results of the daytime energy storage module are combined to determine whether it undertakes power reduction and the amount of power reduction. If the SOC is lower than or equal to the safety lower limit, the daytime energy storage module will no longer participate in further power reduction.

[0078] The fourth step involves selecting the photovoltaic power generation units in order of priority for negligible areas over weakly affected areas, and weakly affected areas over strongly affected areas. When multiple photovoltaic power generation units exist in the same area, the current output power margin of each unit is further compared. Units with larger output power margins are given priority to bear the power reduction, so as to make the power reduction distribution more balanced.

[0079] The fifth step is to allocate the remaining active power deficit to the selected photovoltaic power generation units in sequence. The amount of power reduction undertaken by each unit shall not exceed its current maximum capacity reduction. The total power reduction of the photovoltaic power generation units is obtained by summing the power reduction allocated to each selected unit. If it is still not possible to meet all the active power deficits after traversing all units in the negligible area and the weakly affected area, then units in the strongly affected area are selected in descending order of output power margin to undertake the remaining part until the total active power deficit is completely allocated.

[0080] S244. Allocate the total power reduction to the photovoltaic power generation units participating in the power reduction, sort them by the harmonic voltage content at the photovoltaic output port, and select the photovoltaic power generation units with higher harmonic factors to reduce active power first.

[0081] The following describes step S24 in further detail with reference to specific embodiments.

[0082] The photovoltaic DC power generation system is set to include one daytime energy storage module and four photovoltaic power generation units (PV1, PV2, PV3, PV4). The rated power of the system is 100kW. At the current moment, due to the need for oscillation suppression, the total active power to be reduced is 40kW.

[0083] The current active power output of the daytime energy storage module is detected to be 25kW, which is less than the required reduction of 40kW. The daytime energy storage module will reduce its power output to 0kW, based on its maximum allowable reduction of 25kW. The remaining active power deficit is calculated to be 40kW - 25kW = 15kW.

[0084] Based on a 15kW active power deficit, the amplitude of the probe disturbance was determined to be 1.5kW (10% of the deficit). When the system entered a new steady state after the daytime energy storage module reduced its power, short-duration pulse power disturbances with an amplitude of 1.5kW and a duration of 0.2 seconds were applied sequentially to the daytime energy storage module, PV1, PV2, PV3, and PV4, with an interval of 0.5 seconds between disturbances. The low-voltage DC bus voltage and current waveforms were simultaneously acquired. A matrix beam algorithm was applied to the waveforms before and after the disturbance, extracting a damping ratio of 0.085 before the disturbance. The calculated damping ratios and unit power damping influence coefficients for each unit after the disturbance are shown in Table 1.

[0085] Table 1 Calculation results of damping ratio and unit power damping influence coefficient after disturbance

[0086] The absolute values ​​of the unit power damping influence coefficients of the daytime energy storage module and the four photovoltaic power generation units are ranked as follows: PV4 (0.0100) > PV1 (0.0093) > Daytime module (0.0040) > PV2 (0.0033) > PV3 (0.0013). Setting the first threshold to 0.008 and the second threshold to 0.003, the strong influence areas are PV4 and PV1, the weak influence areas are the daytime module and PV2, and the negligible area is PV3.

[0087] The current state of charge (SOC) of the daytime energy storage module is 45%, which is higher than the safety lower limit of 30%, and it is located in the weak impact zone, thus meeting the conditions for further power reduction. Following the order of priority for negligible areas over weak impact areas, PV3, PV2, and the daytime group are selected to undertake the power reduction in sequence: PV3 has a capacity reduction of 8kW, which is fully undertaken; PV2 has a capacity reduction of 10kW, undertaking 5kW of the remaining 7kW (only a further 7kW reduction is needed to meet the shortfall); the daytime energy storage module, located in the weak impact zone, undertakes the final 2kW. At this point, the total active power shortfall of 15kW is allocated, and the total power reduction of the photovoltaic power generation units is the sum of 8kW from PV3 and 5kW from PV2, i.e., 13kW. Units PV4 and PV1 in the strong impact zone do not participate in this power reduction allocation.

[0088] The total power reduction of 13kW was allocated to PV2 and PV3 participating in the power reduction process. The harmonic voltage content at the output ports of PV2 and PV3 was measured: PV2's harmonic voltage content was 4.2%, and PV3's was 2.8%. Based on harmonic factors, PV2's harmonic factor was higher than PV3's; therefore, PV2 took priority in the power reduction, undertaking 5kW of its capacity reduction, while PV3 undertook the remaining 8kW. At this point, the active power deficit was fully allocated. After the power reduction was implemented, the system damping ratio decreased from 0.085 to 0.078, still within the safe operating range. Simultaneously, PV3, with its lower harmonic content, undertook a larger portion of the power reduction, which helps reduce harmonic injection during the power reduction process.

[0089] It should be noted that when the local low-voltage energy storage (daytime group and nighttime group) does not have the ability to suppress the risk of DC bus oscillation, the DC-connected collaborative control system will activate the collaborative suppression strategy for DC bus risk. After receiving the harmonic suppression risk request, the DC collaborative control system will detect the current margin of the low-voltage energy storage in the DC transformer branch interconnected by the diode isolating switch. When the current margin of the low-voltage energy storage in the DC transformer branch is greater than the reverse harmonic current required by Equation 1, the forward transfer isolating switch will be closed to achieve collaborative harmonic suppression.

[0090] When the oscillation cannot be suppressed by outputting reverse harmonic current through low-voltage energy storage, calculate the required reduction in active power:

[0091] ;

[0092] If the active power output of the DC transformer detected during the daytime low-voltage energy storage exceeds the calculated active power reduction, the active power reduction by the daytime low-voltage energy storage will be used to suppress the problem. The allocation is based on the key indicators of energy storage module temperature and equivalent damping, with the highest temperature as the first priority and the largest equivalent damping as the second priority for power reduction allocation.

[0093] If the active power output of daytime low-voltage energy storage is less than the required reduction in active power, the daytime low-voltage energy storage will be used to reduce the power first, and then the remaining required reduction in active power will be limited by DC power generation units with sufficient sunlight and high active power output. The reduction in harmonic content is as follows:

[0094] ;

[0095] In the formula, The input voltage is represented by D, the duty cycle of the switching transistor is represented by D, and the switching frequency of the converter is represented by f. Indicates photovoltaic filter inductor, To reduce harmonic content.

[0096] The photovoltaic output port is sorted by harmonic voltage content, and the photovoltaics with large harmonic factors (large proportion of harmonic voltage) are given priority to reduce active power.

[0097] S3 and S2 oscillation suppression processes run in parallel. During periods of sufficient sunlight, the DC transformer detects the input current on the low-voltage side. When the current exceeds the limit, it sends a power limit request to the DC collaborative control system and initiates local-level optimization. The daytime group and the nighttime group work together to charge the energy storage. If charging is not possible, the collaborative-level optimization control strategy is initiated.

[0098] S3 includes:

[0099] S31. During periods of sufficient sunlight, the DC transformer monitors the input DC current on the low-voltage side in real time. When the DC current exceeds a preset threshold, the DC transformer sends a power limiting request to the DC collaborative control system.

[0100] S32. After receiving the power limiting request, the DC cooperative control system starts the local-level optimization control strategy of the DC transformer.

[0101] S33. After the local-level optimization control strategy of the DC transformer is started, check whether the night group of the low-voltage DC bus has charging conditions.

[0102] S34. If the nighttime group has charging conditions, the group will be allocated based on the key indicators of energy storage module temperature, health status and equivalent damping, with the optimal temperature as the first priority and the minimum equivalent damping as the second priority.

[0103] S35, DC collaborative control system sends charging power command to night group, night group peak shaving transfer of surplus power of photovoltaic power generation unit, realizes transfer of surplus power of DC transformer to local low voltage energy storage;

[0104] S36. If the night group does not have charging conditions, then further test whether the day group has charging conditions.

[0105] S37. If the daytime group has charging conditions, charging allocation is carried out according to the same priority rules as in step S34; the surplus power of the photovoltaic power generation unit is transferred by the daytime group to realize the transfer of the surplus power of the DC transformer to the local low-voltage energy storage.

[0106] S38. If the local energy storage of both the daytime group and the nighttime group is not available for charging, the DC collaborative control system will activate the collaborative-level optimization control strategy.

[0107] It should be noted that when there is sufficient light, the DC transformer monitors the input DC current on the low-voltage side in real time. When the DC current exceeds the set threshold (preferably 1.05 pu), the DC transformer sends a power limiting request to the DC cooperative control system.

[0108] After receiving a power limiting request from the DC transformer, the DC collaborative control system first activates the local-level optimization control strategy for the DC transformer. With the goal of increasing nighttime power generation duration, it first checks whether the low-voltage DC bus's nighttime energy storage is ready for charging (when SOC < the set threshold, e.g., 85%). If the low-voltage energy storage is ready, it allocates power based on key indicators such as energy storage module temperature, SOH (health status), and equivalent damping, prioritizing optimal temperature and low equivalent damping. The DC collaborative control system then issues a charging power command to the nighttime low-voltage energy storage, which then performs peak shaving and transfers the surplus power from the photovoltaic power generation unit, thus transferring the surplus power of the DC transformer to the local low-voltage energy storage. The specific details are shown in the following formula:

[0109] ;

[0110] In the formula, P total The total power component is given by SOH, the battery pack health status is given by D, the equivalent damping is given by k, the temperature coefficient is given by limit, and the temperature limit is given by limit. Let i be the aging weighting factor for the i-th battery pack. Let i be the time-varying degradation coefficient of the i-th battery pack. Let be the temperature of the i-th battery pack.

[0111] If the energy storage does not have charging conditions at night, further testing is conducted to determine if the low-voltage energy storage during the day has charging conditions (when SOC < the set threshold, such as 85%). If the low-voltage energy storage during the day has charging conditions, the surplus power of the photovoltaic power generation unit is transferred by the low-voltage energy storage during the day for peak shaving. The allocation is based on the key indicators of energy storage module temperature and equivalent damping, with the highest priority being optimal temperature and the lowest equivalent damping being the second priority for charging allocation; thus realizing the transfer of surplus power from the DC transformer to the local low-voltage energy storage.

[0112] S4. After the collaborative optimization control strategy is started, the harmonic content and active damping injection status of each interconnected low-voltage bus are detected, the target branch is screened and cross-branch power transfer is performed; if the target branch can only partially take over, the remaining power is absorbed by the low-voltage energy storage of the interconnected branch. If there is still no effective transfer path, the power is reduced by virtual impedance control. After completion, normal monitoring is entered.

[0113] S4 includes:

[0114] S41. After the DC collaborative control system starts the collaborative-level optimization control strategy, it detects the harmonic content of each interconnected low-voltage DC bus, marks the branches with harmonic content greater than the warning value as not participating in power transfer with priority, then checks whether each branch has implemented active damping, marks the branches with injected damping as not having priority transfer conditions, and finally selects the target branches that meet the conditions and issues power collaborative control reference values.

[0115] S42. After each DC transformer receives the power coordination control reference value, it excludes the marked branches and completes the initial allocation of transferred power based on the DC bus harmonic content margin.

[0116] S43. During the transfer process, the current harmonic content of each branch is detected in real time, the harmonic sensitivity factor is calculated synchronously, the power allocation to the branch with a large harmonic sensitivity factor is suspended, and the power that has not been transferred is re-allocated to the branch with a small harmonic sensitivity factor; at the same time, each DC transformer corrects the voltage reference value based on the virtual impedance, closes the corresponding forward transfer isolating switch, and performs the cross-branch transmission of surplus power.

[0117] S44. If all interconnected DC transformer branches only have partial power transfer capability and cannot take on all surplus power, then the SOC value of the low-voltage energy storage of the interconnected branches shall be further tested.

[0118] S45. If the SOC is lower than the set threshold, it is determined that the charging conditions are met. The charging power command is given priority to the night group. If the night group does not meet the charging conditions, it is switched to the day group. The optimal temperature is the first priority and the minimum equivalent damping is the second priority for charging allocation, and the remaining surplus power is absorbed.

[0119] S46. If there is still no effective power transfer path after the branch-to-branch crossover transfer and interconnected branch-to-branch energy storage absorption, then issue a power limit value to the photovoltaic power generation unit operating in maximum power point tracking mode.

[0120] S47. After receiving the power limit value, the photovoltaic power generation unit starts virtual impedance control, generates a corresponding virtual impedance based on the required power reduction and connects it in series to its own control loop to reduce the output active power.

[0121] It should be noted that, as Figures 2-3 As shown, if the local energy storage does not have charging conditions, the DC collaborative control system activates the collaborative-level optimization control strategy. The DC collaborative control system detects whether the DC power generation units interconnected through diode isolating switches have the conditions for collaborative-level optimization control transfer. When the harmonic content of the interconnected low-voltage DC bus is greater than the warning value, such as 4%, it is determined that the DC transformer has a risk of harmonic amplification and power transfer is not prioritized. It further determines whether the DC transformer has reduced the harmonic content through active damping injection. If it has been actively damped, this DC transformer branch does not have the conditions for priority power transfer.

[0122] If the interconnected DC power generation unit meets the conditions for coordinated optimization transfer, the DC coordinated control system issues a power coordinated control reference value. After receiving the power command, the DC transformer changes the DC transformer voltage reference value through active damping injection, as shown in the following formula:

[0123] ;

[0124] In the formula, V ref，i V is the reference value for the voltage of the i-th DC transformer; nom The rated DC voltage; Rv and sLv are virtual impedances; P out，i Let be the output power value of the i-th DC transformer.

[0125] The DC collaborative control system issues a forward transfer isolation switch for closing power (utilizing the unidirectional conductivity of diodes, such as closing switch 2 when transferring power from DC power generation unit 1 to DC power generation unit 2), and the surplus power of the DC transformer is transferred to the interconnected DC transformer, and the transferred power continues to be transmitted to the AC power grid system;

[0126] If a single DC transformer cannot coordinate and optimize all power, the DC coordinated control system will allocate the power command to be transferred according to the transfer capability of the DC transformer. The DC transformer will calculate the new DC transformer reference value for each DC transformer to achieve power transfer.

[0127] DC transformer branches whose harmonic content has been reduced through active damping are not included in the sorting process; power transfer is allocated based on the DC bus harmonic content margin, and the power transfer is shown in the following formula:

[0128] ;

[0129] In the formula, H represents the priority coefficients for low frequency, subsynchronous, and wideband frequencies; H represents the harmonic content of low frequency, subsynchronous, and wideband frequencies. is the harmonic sensitivity coefficient; P is the current power; k is the harmonic safety factor.

[0130] During the power transfer process, the current harmonic content is detected in real time. When the harmonic content is low, the power transfer is optimized. During the transfer process, the harmonic sensitivity factor is detected in real time (the higher the harmonic content, the greater the harmonic sensitivity when transferring the same power), and the temporary power with high harmonic sensitivity is transferred and allocated. The untransferred power is then redistributed to the DC transformer branch with low harmonic sensitivity factor.

[0131] The DC collaborative control system detects that all interconnected DC transformers only have partial power transfer capability. It further detects whether the low-voltage energy storage has power transfer capability (SOC is less than the threshold, for example, preferably 85%). If the low-voltage energy storage has the remaining power for collaborative optimization, the low-voltage energy storage collaborative optimization control strategy is activated, and the active power is sent to the low-voltage energy storage, which coordinates the power transfer.

[0132] With the goal of increasing nighttime power generation duration, the system first checks whether the low-voltage DC bus has charging conditions at night (when SOC < the set threshold, such as 85%). If the low-voltage energy storage has charging conditions at night, the DC collaborative control system issues a charging power command to the low-voltage energy storage at night. The allocation is based on the key indicators of energy storage module temperature and equivalent damping, with optimal temperature as the first priority and low equivalent damping as the second priority. The surplus power of the photovoltaic power generation unit is transferred by the low-voltage energy storage at night to the peak, thus realizing the transfer of the surplus power of the DC transformer to the local low-voltage energy storage.

[0133] If nighttime energy storage is not available for charging, further testing is conducted to determine if daytime low-voltage energy storage is available for charging (when SOC < the set threshold, such as 85%). If daytime low-voltage energy storage is available for charging, the surplus power of the photovoltaic power generation unit is transferred by the daytime low-voltage energy storage for peak shaving. The allocation is based on key indicators such as the temperature and equivalent damping of the energy storage module, with the highest priority being optimal temperature and the lowest equivalent damping being the second priority for charging allocation; thus realizing the transfer of surplus power from the DC transformer to local low-voltage energy storage.

[0134] When the DC-DC collaborative control system detects a power transfer path without interconnection, it activates the power limiting mode for the photovoltaic (PV) power generation unit (only MPPT with maximum power point tracking activates the power limiting mode). The required power limiting value is sent to the corresponding PV power generation unit. Upon receiving the power limiting value, the PV power generation unit initiates virtual impedance control, generating a virtual impedance based on the required power reduction and connecting it in series with the control loop. Specifically, as follows... Figure 5 As shown, Figure 5 In the middle, PV u Photovoltaic voltage, PV i For photovoltaic current, V p IV is the filter value for the photovoltaic output voltage. u This represents the filtered value of the photovoltaic output current, where Dref is the duty cycle and L is the value of the output current. v and R v This is a virtual impedance.

[0135] S5. When a DC transformer fault is detected during routine monitoring, the low-voltage energy storage is disconnected from the grid to maintain the bus voltage. After the bus stabilizes, power transfer coordination is performed. After the faulty transformer recovers, the power transfer command is withdrawn, and each power generation unit exits the virtual damping distribution and resumes natural power transmission.

[0136] S5 includes:

[0137] S51. When the DC transformer trips due to a fault lockout, the low-voltage energy storage starts the local off-grid control to maintain the low-voltage DC bus voltage stability.

[0138] S52. After receiving the DC transformer blocking signal, the DC cooperative control system continuously monitors the operating status of the low-voltage DC bus. When the bus voltage returns to stability, it starts the power transfer cooperative control mode and executes S4 cross-branch power transfer.

[0139] S53. If, after investigation by S4, it is determined that there is no effective power transfer path, the low-voltage energy storage will continue to maintain the low-voltage DC bus voltage stability. When the low-voltage energy storage SOC rises to the charging limit, a shutdown request will be sent to the DC cooperative control system.

[0140] S54. After receiving the request, the DC collaborative control system first sends a lockout command to the photovoltaic power generation unit. After the photovoltaic power generation unit stops operating, it then sends a shutdown command to the low-voltage energy storage.

[0141] S55. If the faulty branch can transfer the output power of the photovoltaic power generation unit to other normal DC transformer branches through the corresponding diode isolation switch, then after the power transfer is completed and the bus runs smoothly, the low-voltage energy storage will exit the DC voltage control mode and switch to the SOC equalization control mode. The power will be slowly charged and discharged with the goal of setting the optimal SOC. The original faulty DC transformer will remain out of service and await repair.

[0142] S56. When the faulty DC transformer returns to normal operation, the DC cooperative control system retracts all power transfer commands, and each DC transformer participating in the power transfer synchronously exits the virtual damped active power distribution control mode based on power transfer. The power of each generating unit is restored to the normal operating state of being naturally sent out through the DC transformer of this branch.

[0143] It should be noted that when the DC transformer is tripped due to a fault, the low-voltage energy storage detection starts the local off-grid control to maintain the low-voltage DC bus voltage stability. When the DC cooperative control system receives the DC transformer being tripped and further detects that the low-voltage DC bus has recovered to stability, it starts the power transfer cooperative control mode and executes the above step S4.

[0144] If there is no power transfer path in step S4, then the low-voltage energy storage will continuously maintain the stability of the DC bus of the DC power generation unit. When the SOC of the low-voltage energy storage reaches the charging limit, it sends a request to the DC cooperative control to shut down. After receiving the request, the DC cooperative device sends a lockout command to the photovoltaic DC power generation unit. When the photovoltaic shuts down, it sends a shutdown command to the low-voltage energy storage.

[0145] After the photovoltaic power generation units of the faulty branch transformer are interconnected and transferred to the output of other DC transformer branches by means of diode isolating switches, the low-voltage energy storage exits the DC voltage control mode and switches to maintaining SOC balance control; with the optimal SOC as the target, when the SOC is less than the optimal value to 65%, slow charging is started, and the slow charging power is calculated based on the current DC transformer power as 1%; when the SOC is greater than the optimal value to 75%, slow discharging is started, and the slow discharging power is calculated based on the current DC transformer power as 1%, so as not to affect the power characteristics of the DC power generation system to the outside.

[0146] When the DC transformer resumes operation, the DC collaborative control system will retract the power transfer command, and the DC transformer and low-voltage energy storage that transferred the power will exit the active power control mode based on virtual damping distribution of power transfer, so that the power can be output naturally through the DC transformer of this power generation unit.

[0147] According to another embodiment of the present invention, a collaborative optimization control system for a photovoltaic DC power generation system based on virtual damping injection is also provided, the system comprising:

[0148] The oscillation detection and suppression module is used to extract the DC voltage of the low-voltage DC bus from the DC transformer and determine the oscillation risk. When there is a risk, it calculates the required reverse harmonic current and outputs the reverse harmonic current by the day group and the night group. If the total energy storage current margin of the day group and the night group is still insufficient, it sends an oscillation suppression request to the DC collaborative control system.

[0149] The collaborative oscillation suppression module is used by the DC collaborative control system to detect the low-voltage energy storage current margin of the interconnected branches after receiving the oscillation suppression request. If the margin is met, collaborative harmonic suppression is performed. If the margin of all interconnected branches is not met, the required reduction of active power is calculated. After the daytime energy storage reduces the power, the remaining part is limited by the photovoltaic power generation unit according to the harmonic voltage content.

[0150] The local power optimization module is used in parallel with the cooperative oscillation suppression module. The DC transformer detects the low-voltage side input current during periods of sufficient sunlight. When the current exceeds the limit, it sends a power limit request to the DC cooperative control system and starts local optimization. The daytime group and the nighttime group coordinate energy storage charging. When charging is not possible, the cooperative optimization control strategy is activated.

[0151] The collaborative power transfer module is used to detect the harmonic content and active damping injection status of each interconnected low-voltage bus after the collaborative-level optimization control strategy is started, screen the target branch and perform cross-branch power transfer; if the target branch can only partially take over, the remaining power is absorbed by the low-voltage energy storage of the interconnected branch. If there is still no effective transfer path, the power is reduced through virtual impedance control. After completion, it enters normal monitoring.

[0152] The fault condition control module is used to maintain the bus voltage by disconnecting the low-voltage energy storage when a DC transformer fault is detected during routine monitoring. After the bus stabilizes, it performs power transfer coordination. After the faulty transformer recovers, the power transfer command is withdrawn, and each power generation unit exits the virtual damping distribution and resumes natural power transmission.

[0153] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A collaborative optimization control method for a photovoltaic DC power generation system based on virtual damping injection, characterized in that, The method includes: S1. Extract the DC voltage of the low-voltage DC bus from the DC transformer and determine the oscillation risk. If there is a risk, calculate the required reverse harmonic current and output the reverse harmonic current by the day group and the night group. If the total margin of the energy storage current of the day group and the night group is still insufficient, send an oscillation suppression request to the DC collaborative control system. S2. After receiving the oscillation suppression request, the DC collaborative control system detects the low-voltage energy storage current margin of the interconnected branches. If the margin is met, collaborative harmonic suppression is performed. If the margin of all interconnected branches is not met, the required reduction of active power is calculated. After the daytime energy storage reduces the power, the remaining part is limited by the photovoltaic power generation units according to the harmonic voltage content. S3, in parallel with the S2 oscillation suppression process, the DC transformer detects the low-voltage side input current during periods of sufficient sunlight. When the current exceeds the limit, it sends a power limit request to the DC collaborative control system and initiates local-level optimization. The daytime group and the nighttime group work together to charge the energy storage. If charging is not possible, the collaborative-level optimization control strategy is initiated. S4. After the collaborative optimization control strategy is started, the harmonic content and active damping injection status of each interconnected low-voltage bus are detected, the target branch is screened and cross-branch power transfer is performed; if the target branch can only partially take over, the remaining power is absorbed by the low-voltage energy storage of the interconnected branch. If there is still no effective transfer path, the power is reduced by virtual impedance control. After completion, it enters normal monitoring. S5. When a DC transformer fault is detected during routine monitoring, the low-voltage energy storage is disconnected from the grid to maintain the bus voltage. After the bus stabilizes, power transfer coordination is performed. After the faulty transformer recovers, the power transfer command is withdrawn, and each power generation unit exits the virtual damping distribution and resumes natural power transmission.

2. The collaborative optimization control method for a photovoltaic DC power generation system based on virtual damping injection according to claim 1, characterized in that, S1 includes: S11. The DC voltage of the low-voltage DC bus is collected in real time through the DC transformer, and the AC voltage component is extracted through the high-pass filter. Based on the AC voltage component, it is determined whether there is a risk of DC oscillation in the low-voltage DC bus. If so, the required reverse harmonic current is calculated. S12. Check whether the daytime group's energy storage current margin capacity meets the required reverse harmonic current. If it does, allocate based on the key indicators of energy storage module temperature and equivalent internal resistance, with maintaining optimal temperature as the first priority and minimizing equivalent damping as the second priority, and allocate the output harmonic voltage, so that the daytime group outputs the reverse harmonic current. S13. If the energy storage current margin of the daytime group cannot meet the demand, the energy storage current margin of the nighttime group will be further checked. If the total energy storage current margin of the nighttime group and the daytime group meets the demand, the harmonic output command will be allocated according to the same priority rule as S23, and the nighttime group and the daytime group will output the reverse harmonic current in concert. S14. If the total energy storage current margin of the night group and the day group is still insufficient to meet the demand, an oscillation suppression request is sent to the DC collaborative control system to trigger the collaborative suppression of DC bus oscillation risk strategy.

3. The collaborative optimization control method for a photovoltaic DC power generation system based on virtual damping injection according to claim 1, characterized in that, S2 includes: S21. After receiving the oscillation suppression request, the DC cooperative control system detects the current margin of the low-voltage energy storage of the DC transformer branch interconnected by the diode isolating switch. If the current margin meets the required reverse harmonic current output, the forward transfer isolating switch will be closed to achieve cooperative harmonic suppression. S22. If the current margin of all branch low-voltage energy storage cannot meet the required output reverse harmonic current, calculate the required reduction in active power and detect the active power output of daytime group energy storage. S23. If the active power output of the daytime energy storage group meets the derating requirement, the allocation is based on the key indicators of energy storage module temperature and equivalent damping, with the highest temperature as the first priority and the maximum equivalent damping as the second priority, and the power reduction is carried out by the daytime group to reduce the active power. S24. If the active power output of the daytime energy storage group is insufficient to cover all the derated requirements, the active power of the daytime group shall be reduced first, and then the remaining active power to be reduced shall be limited by the DC generation unit with the highest output active power. Among the DC power generation units participating in power limiting, the photovoltaic output port harmonic voltage content is used to sort them, and the photovoltaic units with higher harmonic factors are selected to reduce active power first.

4. The collaborative optimization control method for a photovoltaic DC power generation system based on virtual damping injection according to claim 3, characterized in that, S24 includes: S241. If the active power output by the daytime group is less than the required reduction in active power, the daytime group shall reduce the active power first, and then the remaining active power deficit shall be calculated. S242. Based on the remaining active power deficit, a probing power disturbance is applied sequentially to the daytime group and the photovoltaic power generation unit, and the change in system damping ratio is measured simultaneously. The unit power damping influence coefficient of the daytime group and the photovoltaic power generation unit is calculated based on the change in damping ratio. S243. Based on the unit power damping influence coefficient, the photovoltaic power generation unit is divided into a strong influence zone, a weak influence zone, and a negligible zone. Taking into account the energy storage charge status and the zoning results, the units in the weak influence zone and the negligible zone are given priority to bear the power reduction, and the total power reduction of the photovoltaic power generation unit is determined. S244. Allocate the total power reduction to the photovoltaic power generation units participating in the power reduction, sort them by the harmonic voltage content at the photovoltaic output port, and select the photovoltaic power generation units with higher harmonic factors to reduce active power first.

5. The collaborative optimization control method for a photovoltaic DC power generation system based on virtual damping injection according to claim 4, characterized in that, S242 includes: S2421. Determine the amplitude of the detection disturbance based on the remaining active power deficit, apply short-time pulse power disturbances to the daytime group and photovoltaic power generation unit in sequence, and simultaneously collect the voltage and current waveforms of the low-voltage DC bus. S2422. Preprocess the voltage and current waveforms before and after the disturbance, and use the matrix beam strategy to extract the damping ratio of the dominant oscillation mode of the system. Record the damping ratio before the disturbance and the damping ratio after the disturbance respectively. S2423. Calculate the damping ratio change corresponding to the daytime group and the photovoltaic power generation unit, wherein the damping ratio change is the difference between the damping ratio after the disturbance and the damping ratio before the disturbance. S2424. Based on the change in damping ratio and the corresponding applied disturbance power amplitude, calculate the unit power damping influence coefficient of the daytime group and the photovoltaic power generation unit, respectively.

6. The collaborative optimization control method for a photovoltaic DC power generation system based on virtual damping injection according to claim 5, characterized in that, S2422 includes: S24221. Extract the low-voltage DC bus voltage sampling sequences for the corresponding durations before and after the disturbance, remove the DC bias component, and extract several oscillating AC component sequences as input signals. S24222. Set the total number of sampling points and bundle parameters for each group of input signals, and construct mode identification matrices that differ by one sampling step for each of the oscillating AC component sequences. S24223. Solve for the pseudo-inverse matrix of the preceding mode identification matrix corresponding to each group of input signals, multiply the pseudo-inverse matrix with the following mode identification matrix, solve for all non-zero eigenvalues ​​of the product matrix, and obtain the pole sequence corresponding to each oscillation mode. S24224. Calculate the attenuation factor and oscillation angular frequency of each oscillation mode under the corresponding working condition based on each set of pole sequences. Solve for the complex amplitude of each oscillation mode using the least squares method and select the oscillation mode with the largest complex amplitude as the dominant oscillation mode. S24225. Based on the attenuation factor and oscillation angular frequency of the dominant oscillation mode, calculate and record the damping ratio of the dominant oscillation mode of the system before and after the disturbance.

7. The collaborative optimization control method for a photovoltaic DC power generation system based on virtual damping injection according to claim 1, characterized in that, S3 includes: S31. During periods of sufficient sunlight, the DC transformer monitors the input DC current on the low-voltage side in real time. When the DC current exceeds a preset threshold, the DC transformer sends a power limiting request to the DC collaborative control system. S32. After receiving the power limiting request, the DC cooperative control system starts the local-level optimization control strategy of the DC transformer. S33. After the local-level optimization control strategy of the DC transformer is started, check whether the night group of the low-voltage DC bus has charging conditions. S34. If the nighttime group has charging conditions, the group will be allocated based on the key indicators of energy storage module temperature, health status and equivalent damping, with the optimal temperature as the first priority and the minimum equivalent damping as the second priority. S35, DC collaborative control system sends charging power command to night group, night group peak shaving transfer of surplus power of photovoltaic power generation unit, realizes transfer of surplus power of DC transformer to local low voltage energy storage; S36. If the night group does not have charging conditions, then further test whether the day group has charging conditions. S37. If the daytime group has charging conditions, charging allocation shall be carried out according to the same priority rules as in step S34; the surplus power of the photovoltaic power generation unit shall be transferred by the daytime group to the peak shaving and transfer, so as to transfer the surplus power of the DC transformer to the local low-voltage energy storage. S38. If the local energy storage of both the daytime group and the nighttime group is not available for charging, the DC collaborative control system will activate the collaborative-level optimization control strategy.

8. The collaborative optimization control method for a photovoltaic DC power generation system based on virtual damping injection according to claim 1, characterized in that, S4 includes: S41. After the DC collaborative control system starts the collaborative-level optimization control strategy, it detects the harmonic content of each interconnected low-voltage DC bus, marks the branches with harmonic content greater than the warning value as not participating in power transfer with priority, then checks whether each branch has implemented active damping, marks the branches with injected damping as not having priority transfer conditions, and finally selects the target branches that meet the conditions and issues power collaborative control reference values. S42. After each DC transformer receives the power coordination control reference value, it excludes the marked branches and completes the initial allocation of transferred power based on the DC bus harmonic content margin. S43. During the transfer process, the current harmonic content of each branch is detected in real time, the harmonic sensitivity factor is calculated synchronously, the power allocation to the branch with a large harmonic sensitivity factor is suspended, and the power that has not been transferred is re-allocated to the branch with a small harmonic sensitivity factor; at the same time, each DC transformer corrects the voltage reference value based on the virtual impedance, closes the corresponding forward transfer isolating switch, and performs the cross-branch transmission of surplus power. S44. If all interconnected DC transformer branches only have partial power transfer capability and cannot take on all surplus power, then the SOC value of the low-voltage energy storage of the interconnected branches shall be further tested. S45. If the SOC is lower than the set threshold, it is determined that the charging conditions are met. The charging power command is given priority to the night group. If the night group does not meet the charging conditions, it is switched to the day group. The optimal temperature is the first priority and the minimum equivalent damping is the second priority for charging allocation, and the remaining surplus power is absorbed. S46. If there is still no effective power transfer path after the branch-to-branch crossover transfer and interconnected branch-to-branch energy storage absorption, then issue a power limit value to the photovoltaic power generation unit operating in maximum power point tracking mode. S47. After receiving the power limit value, the photovoltaic power generation unit starts virtual impedance control, generates a corresponding virtual impedance based on the required power reduction and connects it in series to its own control loop to reduce the output active power.

9. The collaborative optimization control method for a photovoltaic DC power generation system based on virtual damping injection according to claim 1, characterized in that, S5 includes: S51. When the DC transformer trips due to a fault lockout, the low-voltage energy storage starts the local off-grid control to maintain the low-voltage DC bus voltage stability. S52. After receiving the DC transformer blocking signal, the DC cooperative control system continuously monitors the operating status of the low-voltage DC bus. When the bus voltage returns to stability, it starts the power transfer cooperative control mode and executes S4 cross-branch power transfer. S53. If, after investigation by S4, it is determined that there is no effective power transfer path, the low-voltage energy storage will continue to maintain the low-voltage DC bus voltage stability. When the low-voltage energy storage SOC rises to the charging limit, a shutdown request will be sent to the DC cooperative control system. S54. After receiving the request, the DC collaborative control system first sends a lockout command to the photovoltaic power generation unit. After the photovoltaic power generation unit stops operating, it then sends a shutdown command to the low-voltage energy storage. S55. If the faulty branch can transfer the output power of the photovoltaic power generation unit to other normal DC transformer branches through the corresponding diode isolation switch, then after the power transfer is completed and the bus runs smoothly, the low-voltage energy storage will exit the DC voltage control mode and switch to the SOC equalization control mode. The power will be slowly charged and discharged with the goal of setting the optimal SOC. The original faulty DC transformer will remain out of service and await repair. S56. When the faulty DC transformer returns to normal operation, the DC cooperative control system retracts all power transfer commands, and each DC transformer participating in the power transfer synchronously exits the virtual damped active power distribution control mode based on power transfer. The power of each generating unit is restored to the normal operating state of being naturally sent out through the DC transformer of this branch.

10. The collaborative optimization control method for a photovoltaic DC power generation system based on virtual damping injection according to claim 2, characterized in that, The formula for calculating the required reverse output current is as follows: ； In the formula, R is the reverse harmonic current; kd is the harmonic damping gain coefficient; V +L V For the required virtual impedance; V thd For the extracted harmonics.

Images