Active power control method based on adaptive droop control

Abstract

The application provides an active power control method based on adaptive droop control, and belongs to the technical field of power transmission control. The application collects operation parameters of each converter station, calculates residual capacity and performs saturation limiting processing; combined with residual capacity, rated capacity and AC system support capability, a comprehensive distribution weight is constructed, and a standby droop coefficient with constraints is calculated in real time. During normal operation, only standby parameters are updated, fault is predicted through electrical quantity sudden change index, and the coefficient is preloaded; when the system is disturbed, the droop coefficient is quickly switched, and then combined with PI control and residual capacity closed loop correction, unbalanced power is distributed according to the carrying capacity of the converter station. The application can avoid overload of heavy load converter station, reduce power sharing pressure of weak AC system, suppress oscillation caused by parameter mutation, shorten fault response time, and improve the operation stability of multi-terminal flexible DC power grid under load fluctuation and equipment fault working condition.

Description

Technical Field

[0001] This invention relates to the field of multi-terminal flexible DC transmission control technology, specifically to an active power control method based on adaptive droop control. Background Technology

[0002] The VSC-MTDC (Variable Reactive Power Grid Conversion) system, with its advantages of independent active and reactive power decoupling control, no commutation failure, and adaptability to new energy grid connection, is widely used in engineering scenarios such as large-scale new energy transmission, inter-regional grid interconnection, and island power supply. Droop control is the core active power distributed distribution strategy of the VSC-MTDC system. By setting a fixed droop coefficient, it achieves automatic power sharing and coordinated DC voltage regulation among multiple converter stations.

[0003] Existing traditional droop control schemes generally adopt a fixed droop coefficient design, where the droop coefficient is set only once based on the rated capacity of the converter station and remains constant throughout the entire operation. This control method has many inherent drawbacks: First, the fixed droop coefficient only considers the rated capacity of the converter station and ignores the real-time load level and remaining operating margin of the converter station. When some converter stations are heavily loaded and the remaining capacity is extremely low, the unbalanced power generated by the fault disturbance will still be evenly distributed to the load, which can easily cause the power of the heavily loaded converter station to exceed the limit and be blocked due to overload, thus expanding the scope of system faults. Second, existing technology does not distinguish between strong and weak AC systems. For weak AC systems with insufficient support capacity, low short-circuit ratio, and poor voltage withstand capability, the disturbance power is still distributed according to a uniform ratio, which can easily cause voltage fluctuations and power oscillations in weak AC systems and reduce the overall stability of the system. Third, traditional control lacks dynamic adaptive adjustment capability. When faced with large disturbances such as sudden load increases or random shutdown of converter stations, the control parameters cannot match the changes in operating conditions in real time, resulting in poor control adaptability. Fourth, conventional adaptive schemes lack parameter limiting, rate of change constraints, and closed-loop correction mechanisms, which can easily lead to problems such as sudden changes in droop coefficient, control oscillation, and power distribution inaccuracy caused by long-term load drift, resulting in poor engineering practicality.

[0004] In summary, existing fixed droop control technology cannot simultaneously take into account the real-time load capacity of the converter station, the differences in AC system support, and the stability under extreme operating conditions. There is an urgent need for an adaptive droop control method that integrates remaining capacity, rated capacity, and AC support capacity. Summary of the Invention

[0005] The technical problem to be solved by this invention is to provide an active power control method based on adaptive droop control, which aims to overcome the shortcomings of existing fixed droop control technology, such as poor adaptability to operating conditions, failure to consider real-time remaining capacity, neglect of differences in AC system strength, susceptibility to overload during faults, and low allocation accuracy during long-term operation. By constructing a multi-dimensional comprehensive allocation weight, parameter amplitude constraint, fault prediction and preloading, and remaining capacity closed-loop feedback correction mechanism, it achieves adaptive and accurate allocation of unbalanced power, where "the greater the capacity, the more it bears," effectively avoiding overload of heavily loaded converter stations, improving the operational stability of weak AC systems, and adapting to various disturbances and extreme operating conditions.

[0006] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: The active power control method based on adaptive droop control includes the following steps: S1. Real-time acquisition of operating parameters of each converter station in the multi-terminal flexible DC transmission system. The operating parameters include the rated active capacity, current load power, AC side electrical parameters, and DC node voltage of each converter station. S2. Based on the rated active capacity and current load power of each converter station, calculate the remaining capacity of each converter station, and perform normalization and saturation limiting on the remaining capacity to obtain the normalized remaining capacity. S3. Combining the normalized remaining capacity, the normalized coefficient of the rated capacity of the converter station, and the AC system support capacity index, the comprehensive allocation weight of each converter station is calculated. S4. Calculate the standby adaptive droop coefficient based on the comprehensive allocation weight and the basic nominal droop coefficient, and set upper and lower limit constraints and adjacent sampling period change rate constraints for the standby adaptive droop coefficient. S5. When the system is running in normal steady state, the backup adaptive droop coefficient is continuously updated and stored in real time, without replacing the current operating droop coefficient of the controller. S6. Real-time monitoring of system fault prediction indicators. When the fault prediction indicators reach the warning threshold, the latest stored backup adaptive droop coefficient is preloaded. S7. When a converter station shutdown, load surge, or power limit violation disturbance event is detected, the current droop coefficient in the controller will be quickly switched to the pre-loaded standby adaptive droop coefficient. S8. The unbalanced power distribution is adjusted by using a PI controller combined with the remaining capacity closed-loop feedback correction circuit, so that the converter station with a higher comprehensive distribution weight can bear more unbalanced power.

[0007] In step S1 above, the AC side electrical parameters and DC node voltage are collected in real time through high-precision voltage transformers and current transformers.

[0008] In step S2 above, the calculation cycle of the remaining capacity is kept consistent with the collection cycle of the operating parameters to ensure that the remaining capacity data is updated synchronously in real time.

[0009] The formula for calculating the remaining capacity in step S2 above is: ; in, Let be the remaining capacity of the j-th converter station. Let j be the rated power of the j-th converter station. Let the current load power of the j-th converter station be denoted as ; the remaining capacity is processed using a saturation limiting function, and a lower limit protection value for the remaining capacity is set. When the remaining capacity is lower than Time uniform This avoids a sudden change in the droop coefficient when the remaining capacity approaches 0.

[0010] The formula for calculating the comprehensive allocation weight in step S3 above is: ; In the formula This represents the comprehensive allocation weight of the j-th converter station at sampling time k. This represents the remaining capacity after normalized clipping. This represents the maximum rated capacity among all converter stations. The dynamic weights for strong and weak AC systems are represented by a, b, and c, which are positive sensitivity indices corresponding to the adjustment weights of remaining capacity, rated capacity, and AC support capacity, respectively.

[0011] The dynamic weights of the aforementioned strong and weak communication system The method of obtaining it is: The AC system support capacity index is calculated by collecting the AC side short-circuit ratio, bus voltage deviation, and current margin of the converter station, and then normalizing and limiting the support capacity index to obtain... The higher the short-circuit ratio of the AC system, the larger the corresponding dynamic weight value.

[0012] The formula for calculating the standby adaptive droop coefficient in step S4 above is: ; In the formula, This represents the standby adaptive droop coefficient calculated at sampling time k for the j-th converter station; Indicates the base sag coefficient; and These are the dynamic lower limit and dynamic upper limit of the droop coefficient for the j-th converter station, respectively; To prevent zero constant; In step S6 above, the fault prediction indicators include AC voltage change rate, DC voltage fluctuation amplitude, and converter station output current change amplitude. If any one of the indicators exceeds the preset threshold, it is determined that the warning threshold has been reached, and the standby droop coefficient preloading operation is executed.

[0013] The disturbance events in step S7 above include converter station short circuit faults, planned / unplanned shutdowns of converter stations, DC line open circuits, and sudden increases in AC side load. The overall switching response time of the droop coefficient is no more than 10ms.

[0014] The remaining capacity closed-loop feedback correction stage in step S8 above is implemented as follows: The system collects the actual unbalanced power shared by each converter station in real time, compares it with the target power shared based on the comprehensive allocation weight, and adds the power difference as a correction amount to the standby adaptive droop coefficient to eliminate the allocation deviation caused by long-term load drift.

[0015] The above-mentioned PI controller proportional gain The integral gain has a value range of 0.5 to 2.0. The value range is 0.1~0.5; the output of the PI controller, together with the remaining capacity correction, adjusts the active power output of each converter station.

[0016] The aforementioned multi-terminal flexible DC transmission system is a four-terminal VSC-HVDC topology, which includes at least two strong AC system-side converter stations and at least two weak AC system-side converter stations; the short-circuit ratio of the strong AC system is greater than a preset threshold, and the short-circuit ratio of the weak AC system is lower than a preset threshold.

[0017] The sensitivity indices a, b, and c mentioned above all range from 0.2 to 3.0. When it is necessary to strengthen the constraint effect of remaining capacity on power allocation, the value of index a should be increased; when it is necessary to weaken the impact of AC system support differences, the value of index c should be decreased.

[0018] The active power control method based on adaptive droop control mentioned in this invention has the following beneficial effects: 1. This invention constructs a multi-dimensional comprehensive weight allocation system that combines remaining capacity, rated capacity, and AC system support capabilities. It breaks through the limitations of traditional power allocation based solely on rated capacity, and achieves precise matching between unbalanced power and the real-time carrying capacity of the converter station. This fundamentally avoids overload and overlimit issues in heavily loaded converter stations and significantly improves the system's fault disturbance tolerance.

[0019] 2. This invention adds multiple stability mechanisms such as normalized limit, lower limit protection, and rate of change constraint, which solves the industry pain points of droop coefficient sudden change and system oscillation when the remaining capacity is close to zero, the load fluctuates drastically, and the stability of operation under extreme conditions is significantly improved.

[0020] 3. This invention pioneers a fault prediction + backup coefficient preloading mechanism, which differs from the traditional passive calculation and switching mode after a fault. It realizes the early preprocessing of disturbances, greatly shortens the control response delay, and improves the dynamic response speed of the system.

[0021] 4. This invention adds a closed-loop feedback correction for remaining capacity to solve the problem of power distribution inaccuracy caused by long-term operating condition drift and ensure the distribution accuracy of the system during long-term steady-state operation.

[0022] 5. This invention quantitatively distinguishes the support differences between strong and weak AC systems, and through dynamic weighting, allows the strong AC system to bear more disturbance power and the weak AC system to share less pressure, effectively protecting the stable operation of the weak AC power grid and adapting to the complex flexible DC topology where multiple strong and weak power grids coexist. Attached Figure Description

[0023] The present invention will be further described below with reference to the accompanying drawings and embodiments: Figure 1 This is a block diagram of the adaptive active power control scheme of the present invention in a converter station; Figure 2 This is a topology diagram of a four-terminal VSC-HVDC system in an embodiment of the present invention; Figure 3 The active power change curves of each converter station under the scenario of sudden load increase; Figure 4 The curves showing the change in active power of each converter station under the scenario of converter station failure and shutdown. Detailed Implementation

[0024] The technical solutions in the embodiments of this application will be clearly and completely described below with reference to the embodiments of this application. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0025] The active power control method based on adaptive droop control includes the following steps: S1. Real-time acquisition of operating parameters of each converter station in the multi-terminal flexible DC transmission system. The operating parameters include the rated active capacity, current load power, AC side electrical parameters, and DC node voltage of each converter station. S2. Based on the rated active capacity and current load power of each converter station, calculate the remaining capacity of each converter station, and perform normalization and saturation limiting on the remaining capacity to obtain the normalized remaining capacity. S3. Combining the normalized remaining capacity, the normalized coefficient of the rated capacity of the converter station, and the AC system support capacity index, the comprehensive allocation weight of each converter station is calculated. S4. Calculate the standby adaptive droop coefficient based on the comprehensive allocation weight and the basic nominal droop coefficient, and set upper and lower limit constraints and adjacent sampling period change rate constraints for the standby adaptive droop coefficient. S5. When the system is running in normal steady state, the backup adaptive droop coefficient is continuously updated and stored in real time, without replacing the current operating droop coefficient of the controller. S6. Real-time monitoring of system fault prediction indicators. When the fault prediction indicators reach the warning threshold, the latest stored backup adaptive droop coefficient is preloaded. S7. When a converter station shutdown, load surge, or power limit violation disturbance event is detected, the current droop coefficient in the controller will be quickly switched to the pre-loaded standby adaptive droop coefficient. S8. The unbalanced power distribution is adjusted by using a PI controller combined with the remaining capacity closed-loop feedback correction circuit, so that the converter station with a higher comprehensive distribution weight can bear more unbalanced power.

[0026] In step S1 above, the AC side electrical parameters and DC node voltage are collected in real time through high-precision voltage transformers and current transformers.

[0027] In step S2 above, the calculation cycle of the remaining capacity is kept consistent with the collection cycle of the operating parameters to ensure that the remaining capacity data is updated synchronously in real time.

[0028] The formula for calculating the remaining capacity in step S2 above is: ; in, Let be the remaining capacity of the j-th converter station. Let j be the rated power of the j-th converter station. Let the current load power of the j-th converter station be denoted as ; the remaining capacity is processed using a saturation limiting function, and a lower limit protection value for the remaining capacity is set. When the remaining capacity is lower than Time uniform This avoids a sudden change in the droop coefficient when the remaining capacity approaches 0.

[0029] The formula for calculating the comprehensive allocation weight in step S3 above is: ; In the formula This represents the comprehensive allocation weight of the j-th converter station at sampling time k. This represents the remaining capacity after normalized clipping. This represents the maximum rated capacity among all converter stations. The dynamic weights for strong and weak AC systems are represented by a, b, and c, which are positive sensitivity indices corresponding to the adjustment weights of remaining capacity, rated capacity, and AC support capacity, respectively.

[0030] The dynamic weights of the aforementioned strong and weak communication system The method of obtaining it is: The AC system support capacity index is calculated by collecting the AC side short-circuit ratio, bus voltage deviation, and current margin of the converter station, and then normalizing and limiting the support capacity index to obtain... The higher the short-circuit ratio of the AC system, the larger the corresponding dynamic weight value.

[0031] The formula for calculating the standby adaptive droop coefficient in step S4 above is: ; In the formula, This represents the standby adaptive droop coefficient calculated at sampling time k for the j-th converter station; Indicates the base sag coefficient; and These are the dynamic lower limit and dynamic upper limit of the droop coefficient for the j-th converter station, respectively; To prevent zero constant; In step S6 above, the fault prediction indicators include AC voltage change rate, DC voltage fluctuation amplitude, and converter station output current change amplitude. If any one of the indicators exceeds the preset threshold, it is determined that the warning threshold has been reached, and the standby droop coefficient preloading operation is executed.

[0032] The disturbance events in step S7 above include converter station short circuit faults, planned / unplanned shutdowns of converter stations, DC line open circuits, and sudden increases in AC side load. The overall switching response time of the droop coefficient is no more than 10ms.

[0033] The remaining capacity closed-loop feedback correction stage in step S8 above is implemented as follows: The system collects the actual unbalanced power shared by each converter station in real time, compares it with the target power shared based on the comprehensive allocation weight, and adds the power difference as a correction amount to the standby adaptive droop coefficient to eliminate the allocation deviation caused by long-term load drift.

[0034] The above-mentioned PI controller proportional gain The integral gain has a value range of 0.5 to 2.0. The value range is 0.1~0.5; the output of the PI controller, together with the remaining capacity correction, adjusts the active power output of each converter station.

[0035] The aforementioned multi-terminal flexible DC transmission system is a four-terminal VSC-HVDC topology, which includes at least two strong AC system-side converter stations and at least two weak AC system-side converter stations; the short-circuit ratio of the strong AC system is greater than a preset threshold, and the short-circuit ratio of the weak AC system is lower than a preset threshold.

[0036] The sensitivity indices a, b, and c mentioned above all range from 0.2 to 3.0. When it is necessary to strengthen the constraint effect of remaining capacity on power allocation, the value of index a should be increased; when it is necessary to weaken the impact of AC system support differences, the value of index c should be decreased.

[0037] This invention breaks away from the traditional control mode of evenly distributing unbalanced power with a fixed droop coefficient. First, it collects electrical operation data from each converter station, calculates the real-time remaining capacity, and adds saturation limiting processing to prevent parameter abrupt changes. Second, it integrates the remaining capacity, the rated capacity of the converter station, and the strong and weak support indicators of the AC system to construct a multi-factor multiplication comprehensive allocation weight; it then uses the inverse proportional relationship between the weight and the droop coefficient to calculate the standby droop coefficient in real time, while simultaneously adding amplitude and rate-of-change constraints to the coefficient.

[0038] Under normal operating conditions, the system continuously updates the reserve coefficient without initiating control. Faults are predicted based on voltage and current surge indicators, and parameters are pre-stored. In the event of disturbances such as converter station outages, line faults, or sudden load increases, the system quickly switches to a pre-loaded adaptive droop coefficient. Finally, a residual capacity feedback correction is added to the PI closed-loop system to allocate unbalanced power according to the comprehensive carrying capacity of each converter station, ensuring that converter stations with higher residual margins and stronger AC grid support share more disturbance power.

[0039] Example 1: I. Adaptive droop control principle in this embodiment In traditional active power control, the droop factor typically depends only on the individual converter ratings. Therefore, for an MTDC grid with proportional converter stations, the overflow power after a converter failure will be distributed evenly. However, under specific operating conditions, the loads on all converter stations may not be equal, and some converter stations may not participate in power distribution fairly. Therefore, the selectable converters should participate in power distribution based not only on their rated power but also on their remaining capacity, i.e., the difference between their rated capacity and the current load.

[0040] To solve this problem, such as Figure 1 As shown, a method for calculating the droop factor based on the remaining capacity is proposed.

[0041] Figure 1 middle Let j be the actual power of the j-th converter station in the PCC. Let the reference power of the j-th converter station be the power in the PCC. This is the voltage measurement value at this node. This is the reference voltage value for that node. Let j be the rated power of the j-th converter station. This is the maximum rated power among all converter stations. For the defined positive constant, The droop coefficient is... It is the nominal droop, which is inversely proportional to the converter station's rated value. PI controller, proportional-integral controller For proportional control gain, This is the integral control gain. During normal operation, the system calculates the remaining capacity of the j-th converter station in real time using the actual power of the j-th converter station in the PCC and the rated power of the j-th converter station, and calculates the corresponding droop factor accordingly. This is a backup, but it will not change the droop coefficient value in the controller. Once the converter station fault detection device detects a fault, the trigger is activated, and the converter station's droop coefficient... It will be immediately replaced with the droop coefficient value calculated before the fault, ensuring that the converter station can allocate load based on the remaining capacity after the fault occurs.

[0042] To ensure consistency between the newly added dynamic control formula and the original steady-state derivation formula, the following notation conventions are adopted in this paper: variables with sampling time k represent instantaneous quantities calculated or acquired by the controller in real time during the k-th sampling period; in post-fault steady-state derivations, original formulas, or textual descriptions that do not emphasize the sampling time, variables with omitted (k) represent the value of the same physical quantity at the current analysis time or steady-state operating point. Therefore, and For the same active power variable at the j-th converter station, the only difference is... Emphasis is placed on real-time sampling. This is a shorthand for the current time or steady-state value; similarly, , When the sampling time is not emphasized, the isoband(k) variable can be simplified as follows: , .

[0043] When a fault occurs in a power system, unbalanced power is generated, which needs to be absorbed by other normal converter stations. Traditional methods of distributing unbalanced power are average distribution or distribution according to a fixed droop coefficient. This invention takes the remaining capacity of the converter station as the core and combines it with the rated capacity and AC system support capacity to form a comprehensive distribution weight. Unbalanced power is distributed according to the comprehensive distribution weight; the larger the comprehensive distribution weight, the smaller the corresponding droop coefficient, and the more unbalanced power is absorbed after a fault.

[0044] This method establishes a relationship between remaining capacity, comprehensive allocation weight, and droop coefficient: the larger the remaining capacity, the greater the power margin that the converter station can continue to bear; the higher the rated capacity, the stronger the station's basic carrying capacity; and the stronger the AC system's support capacity, the stronger the station's ability to withstand new power disturbances after a fault. This application converts the above factors together into a comprehensive allocation weight, and makes the comprehensive allocation weight inversely related to the droop coefficient, that is, the larger the comprehensive allocation weight, the smaller the droop coefficient, and the more unbalanced power it bears after a fault. The original droop correction relationship based on rated capacity and remaining capacity can be used to explain the basic inverse relationship between remaining capacity and droop coefficient; on this basis, this application further introduces normalization, rated capacity weighting, dynamic weights for strong and weak AC systems, amplitude limiting, and rate of change constraints, so that power allocation no longer depends solely on a single remaining capacity, but reflects the real-time comprehensive carrying capacity of the converter station.

[0045] To further enhance the feasibility of the above relationship under extreme operating conditions, this application extends the inverse relationship of remaining capacity to a droop coefficient calculation chain of "normalized remaining capacity, rated power weighting, dynamic weighting of strong and weak AC systems, and dynamic limiting". This extension does not simply change the proportional coefficient, but establishes a coupling relationship between the converter station's operating margin, the AC system's support capacity, and control stability.

[0046] Define the normalized remaining capacity of the j-th converter station at sampling time k as: (1) Where j represents the converter station number, and k represents the sampling time. This represents the rated active capacity of the j-th converter station, and its meaning is consistent with that of the rated power of the j-th converter station in the original formula; This represents the actual active power of the j-th converter station at sampling time k. When the sampling time is not emphasized, or in the steady-state derivation after a fault, it can be simplified to... , compared with the original formula They are the same physical quantity; Used to characterize the current occupied capacity of the converter station; This represents the normalized remaining capacity. The larger the value, the greater the active power margin that the converter station can continue to handle. This represents the saturation limiting function, which is applied when the input value is below... Time to take When the value is higher than 1, take 1; when the value is within the range, take the input value itself. This is the lower limit protection value for remaining capacity, used to prevent the droop coefficient from increasing unbounded when the remaining capacity approaches zero. The comprehensive allocation weight for the j-th converter station is defined as: (2) in, This represents the comprehensive allocation weight of the j-th converter station at sampling time k, which is used to determine the proportion of unbalanced power that the station should bear after a fault. This represents the maximum rated capacity among all converter stations, used for normalizing the rated capacity; a, b, and c are the weighting indices for the remaining capacity, rated capacity, and dynamic weights of the strong and weak AC systems, respectively, used to adjust the influence of these three factors in power allocation; a, b, and c are sensitivity indices rather than linear weighting coefficients. Because... , and All are normalized dimensionless quantities. This application adopts a multiplicative comprehensive weight calculation method, so that if any factor is low, the comprehensive allocation weight will be reduced, thereby avoiding converter stations with insufficient remaining capacity or weak AC system support capacity from bearing too much unbalanced power due to other factors being high. a, b, and c are used to adjust the sensitivity of the comprehensive allocation weight to remaining capacity, rated capacity, and AC system support capacity, respectively. a, b, and c are all positive numbers and can be preset during the system initialization stage; when a=b=c=1, each factor is multiplied and merged according to the original normalized ratio; when it is necessary to strengthen the constraint effect of a certain factor, the corresponding index is increased. This represents the dynamic weighting of the strong and weak AC systems; a larger value indicates stronger AC-side support capability for the j-th converter station. The standby adaptive droop coefficient is calculated as follows: (3) in, This represents the standby adaptive droop coefficient calculated at sampling time k for the j-th converter station, which is used to quickly replace the current droop coefficient when a fault or disturbance is triggered. This represents the base sag coefficient, used to provide a reference level for sag adjustment; and These are the dynamic lower limit and dynamic upper limit of the droop coefficient of the j-th converter station, respectively, used to prevent the control quantity from being too small or too large; To prevent zero constant, used to avoid When the denominator is close to zero, it becomes too small; This is a sensitivity index used to adjust the intensity of the influence of the overall weighting on the droop coefficient. Because... and It is negatively correlated. The larger, The smaller the value, the more unbalanced power the converter station will handle after a fault. To avoid abrupt changes in the droop coefficient, a rate of change constraint is set: (4) in, This represents the maximum allowable variation of the reserve droop coefficient of the j-th converter station within adjacent sampling periods. It is used to limit the rate of change of the droop coefficient, preventing power oscillations caused by sudden parameter changes in the controller. After the above limiting and rate of change constraints, the controller obtains a preloadable reserve droop coefficient. .

[0047] Final process: The system calculates the remaining capacity, the normalized rated capacity, and the dynamic weights of the strong and weak AC systems in real time to form a comprehensive allocation weight, and calculates the standby droop coefficient in real time based on this. During normal operation, the standby droop coefficient is only stored and does not update the current droop coefficient in the controller. When unbalanced power occurs or the fault prediction indicators reach the action conditions, the system quickly replaces the current droop coefficient with the preloaded standby droop coefficient.

[0048] This invention uses the remaining capacity of the converter station as the core basis for adaptively adjusting the droop coefficient, and combines the rated capacity of the converter station and the support capacity of the AC system to form a comprehensive allocation weight. When a node in the system fails and loses power, the additional power borne by other converter stations is related to the comprehensive allocation weight; converter stations with larger remaining capacity, higher rated capacity, and stronger AC systems have a larger comprehensive allocation weight and bear more unbalanced power; converter stations with smaller remaining capacity or weaker AC systems have a smaller comprehensive allocation weight and bear less unbalanced power.

[0049] II. Dynamically adjust the droop coefficient of the converter station based on its remaining capacity: Remaining capacity is defined as the difference between rated power and current load. , (5) In the formula The remaining capacity of converter station j The rated power of converter station j, This represents the current load of converter station j.

[0050] (6) In the formula The droop coefficient is... It is the nominal droop, which is inversely proportional to the converter station's rated value. This is the maximum rated power among all converter stations. Let be a defined positive constant. Suppose that in an n-node network, the converter station at the nth node fails, and let the sum of the powers of all converter stations before the failure be . The unbalanced power generated during a fault is : The solution formula is shown in equation (7): (7) The solution process is shown below. By combining equations (6), (8), (9), (10), and (11), we can obtain the solution. : (8) (9) (10) (11) Before a fault occurs, each converter station calculates the adaptive droop coefficient in real time according to formulas (5) and (6). However, the droop factor in the controller is not updated. When a fault occurs, the droop factor value in the controller is updated immediately. At this time, the droop factor includes the remaining capacity and load level information.

[0051] Determining the dynamic weights of strong and weak communication systems: Real-time acquisition of AC side short-circuit ratio at each converter station AC bus voltage deviation Current margin and converter station availability status And calculate the AC system support capability index: (12) right After normalization and amplitude limiting, we get: in, This represents the AC-side support capacity index of the j-th converter station, used to evaluate the station's AC system's ability to withstand new active power disturbances. This represents the AC side short-circuit ratio of the j-th converter station. This represents the short-circuit ratio reference value. The larger the value, the stronger the communication system. This represents the current margin of the j-th converter station. This indicates the rated current; a larger current margin means the converter station is further away from the current limit. Indicates the AC bus voltage deviation. This indicates the AC voltage reference value. The larger the voltage deviation, the less suitable it is for the AC side to continue to increase power support. Indicates the available status variables of the converter station; , , , These are the weighting coefficients for each indicator, used to adjust the proportion of short-circuit ratio, current margin, voltage deviation, and equipment availability in the AC support capability evaluation.

[0052] (13) in, and These represent the lower and upper limits of the dynamic weights of the strong and weak communication systems, respectively, to prevent the weights of the communication systems from being too small or too large; This represents the sum of the AC-side support capacity indicators of each converter station participating in power distribution, used for... Perform normalization. Substitute into the overall allocation weight This results in converter stations with large remaining capacity and strong AC system support capabilities having smaller droop coefficients, allowing them to handle more unbalanced power after a fault; while converter stations with weak AC systems or small remaining capacity have larger droop coefficients, allowing them to handle less unbalanced power after a fault.

[0053] The post-fault operating point of the MTDC network containing converter stations is analyzed.

[0054] First, according to Figure 1 You can get ;(14) because (15) If we consider this DC network as an equivalent intermediate node that does not generate active power, then we have: (16) in It is the total network loss in a DC network.

[0055] Equivalent transformation of equation (8) yields (17) And thus obtain (18) From equations (17) and (18), we can obtain: (19) (20) Suppose that in an n-node network, the converter station of the nth node fails. ;(twenty one) In this network, the total DC network loss after a power outage It can be represented as ;(twenty two) From equation (18), we can obtain: ;(twenty three) From equations (21), (22), and (23), we can obtain equations (24) and (25), and then obtain the power of the j-th converter station after the fault.

[0056] ;(twenty four) (25) From equations (20) and (25), we can obtain the additional power that the j-th node converter station will bear after the fault. for (26) because and Since the terms in parentheses in equation (26) are very small, they can be ignored, resulting in equation (27): (27) From both sides of equation (27) Differentiation yields (28) From equation (6), it can be seen that increasing the remaining capacity will decrease the droop coefficient. Further combining this with the comprehensive allocation weight, it can be seen that when the rated capacity is higher and the AC system support capability is stronger, the comprehensive allocation weight of the corresponding converter station further increases, and the standby droop coefficient further decreases. From equation (28), it can be seen that when the droop coefficient decreases, the power borne by the j-th node converter station increases. In summary, node converter stations with large remaining capacity, high rated capacity, and strong AC system support capability bear more unbalanced power after a fault occurs.

[0057] Compared to traditional systems with a fixed sag coefficient, there are: (29) From equations (27) and (29), we can obtain: (30) Specifically, when each converter station has the same rated capacity, we have: (31) From equations (30) and (31), we can obtain that when a fault occurs, the additional power borne by the j-th node converter station is... If the value is fixed, it may cause power to exceed the limit for converter stations with small remaining capacity, thereby expanding the scope of the fault.

[0058] In summary, it can be seen that the adaptive active power control method can reasonably allocate the load when the power grid fails or the converter station fails, thus avoiding overload and effectively improving the stability of the power grid.

[0059] III. Fault prediction and backup droop coefficient preloading: Real-time calculation of fault prediction indicators for each converter station: (32) in, This represents the fault prediction index of the j-th converter station at sampling time k, used to determine whether the backup droop coefficient needs to be preloaded in advance; This represents the rate of change of DC voltage at the j-th converter station. The normalized threshold value represents the criterion for DC voltage sudden change. This represents the rate of change of current at the j-th converter station. The normalized threshold represents the current surge criterion. This represents the current active power of the j-th converter station. This indicates the active power limit of the station. Used to characterize the degree of power exceeding the limit; For converter station protection, interlocking, communication abnormality, or out-of-operation status; , , , These are the weighting coefficients for each component in the fault prediction index. (Definition) As the warning threshold, Action threshold, For the preload duration. When Exceeding the warning threshold continuously And continue At this time, the system enters a preloading state and loads the latest... Write to the controller's spare area but do not participate in main control for the time being; when Exceeding the action threshold If a converter station outage, a sudden load surge, or a power limit violation is detected, the trigger immediately switches the current droop factor to a preloaded droop factor. Through the above prediction and preload mechanism, a complete recalculation process is no longer required when a fault occurs, enabling zero-delay or near-zero-delay switching.

[0060] IV. Remaining capacity closed-loop feedback correction: The system calculates the target unbalanced power sharing ratio and target power sharing of the j-th converter station based on the comprehensive allocation weight: (33) (34) in, This represents the target unbalanced power sharing ratio of the j-th converter station. This represents the sum of the overall allocation weights of each converter station participating in power allocation; This represents the unbalanced power generated by the system at sampling time k; This represents the target additional power contribution calculated by the j-th converter station according to the comprehensive allocation weight. The target and actual power contributions with (k) can be omitted in subsequent steady-state analysis or simulation results and denoted as... , This target value is used to provide a benchmark for subsequent feedback correction.

[0061] The actual additional power taken up by the j-th converter station after the fault was collected. This generates an allocation error and feeds it back to the droop coefficient correction stage: (35) (36) in, This indicates the actual increase in power handled by the j-th converter station. Increased power to meet the target Allocation error between them; This indicates the droop coefficient used in the next sampling period; and These represent the proportional coefficient and integral coefficient of the remaining capacity feedback correction circuit, respectively. This represents the accumulated power allocation error. This feedback term is used to correct the actual power sharing result in the droop coefficient: when the actual power received is higher than the target value, the correction stage increases the corresponding droop coefficient, reducing the subsequent power sharing of the converter station; when the actual power received is lower than the target value, the correction stage decreases the corresponding droop coefficient, increasing the subsequent power sharing of the converter station. An integral term is set with a limit to prevent oscillations caused by accumulated errors during long-term operation.

[0062] V. Simulation using PSCAD / EMTDC: This method was established on PSCAD / EMTDC. Figure 2 The effectiveness of the proposed adaptive active power control method is verified using a 4-terminal VSC-HVDC system model shown in the figure. , , , For four communication systems, , , , There are four converter stations.

[0063] As shown in Table 1, S is the line voltage of the AC system, and S is the capacity of the AC system. This is the active power command value for VSC. This is the GSVSC DC voltage command value. The DC voltage droop factor of the converter station is 1.2Ω, and the line resistance between interconnecting converter stations is 1.2Ω.

[0064] Table 1 System Parameters

[0065] To investigate the improvement of active power allocation by adaptive active power control methods, a large power disturbance scenario is selected for simulation. This scenario primarily discusses two situations: a sudden load increase in the system and a converter station out of operation due to a fault. P1-P4 represent the active power transmitted by converter stations 1-4, respectively. Table 1 shows that the AC system... , The capacity S is relatively large, making it a strong AC system. , The capacity S is relatively small, making it a weak AC system.

[0066] Figure 3 The active power change curves of each converter station when the load of AC system 1 suddenly increases by 300MW at t=15s, using the adaptive active power control method to adjust the unbalanced power distribution ratio.

[0067] Depend on Figure 3 It is known that ΔP1 is -183.17 MW, ΔP2 is 34.88 MW, ΔP3 is 111.95 MW, and ΔP4 is 35.58 MW. The adaptive active power allocation method redistributes unbalanced power by dynamically adjusting the droop coefficient. Converter station 3 has a larger remaining capacity and stronger AC system support capability, therefore its overall allocation weight is larger, allowing it to support more power. Converter stations 2 and 4 have smaller remaining capacities or weaker AC system support capabilities, therefore their overall allocation weight is smaller, allowing them to share less unbalanced power, demonstrating the adaptive unbalanced power allocation characteristics based on the overall allocation weight. This significantly improves the stable operation capability of the weak AC system while ensuring the stability of the strong AC system. Simulation results show that after a sudden load increase, the adaptive active power control method, which comprehensively considers the remaining capacity, rated capacity, and AC system support capability, can effectively improve the power allocation of the weak AC system compared to traditional droop control.

[0068] When using fixed coefficient droop control, ΔP2, ΔP3, and ΔP4 are 132.83 MW, 166.56 MW, and 99.65 MW, respectively. After exiting the program, If the unbalanced power is too much and reaches the power limit, it seriously threatens the stable operation of the MTDC system.

[0069] After using the adaptive active power allocation method, ΔP2 decreased to 68.66 MW, ΔP3 became 237.34 MW, and ΔP4 became 92.73 MW. The transmitted active power was significantly lower than the power limit, with the strong AC system 3 bearing more of the unbalanced power. Simulation waveforms revealed that when the converter station power deviated significantly from the reference value, the adaptive active power allocation method could effectively prevent the converter station's transmitted power from reaching the power limit, thereby improving the VSC operating conditions.

[0070] The adaptive active power allocation method can dynamically adjust the proportion of power increments shared by the converter station based on the remaining capacity of the VSC, so as to preserve the regulation capacity of the converter station as much as possible and ensure that the VSC-MTDC system operates at a relatively stable level.

Claims

1. An active power control method based on adaptive droop control, characterized in that, Includes the following steps: S1. Real-time acquisition of operating parameters of each converter station in the multi-terminal flexible DC transmission system. The operating parameters include the rated active capacity, current load power, AC side electrical parameters, and DC node voltage of each converter station. S2. Based on the rated active capacity and current load power of each converter station, calculate the remaining capacity of each converter station, and perform normalization and saturation limiting on the remaining capacity to obtain the normalized remaining capacity. S3. Combining the normalized remaining capacity, the normalized coefficient of the rated capacity of the converter station, and the AC system support capacity index, the comprehensive allocation weight of each converter station is calculated. S4. Calculate the standby adaptive droop coefficient based on the comprehensive allocation weight and the basic nominal droop coefficient, and set upper and lower limit constraints and adjacent sampling period change rate constraints for the standby adaptive droop coefficient. S5. When the system is running in normal steady state, the backup adaptive droop coefficient is continuously updated and stored in real time, without replacing the current operating droop coefficient of the controller. S6. Real-time monitoring of system fault prediction indicators. When the fault prediction indicators reach the warning threshold, the latest stored backup adaptive droop coefficient is preloaded. S7. When a converter station shutdown, load surge, or power limit violation disturbance event is detected, the current droop coefficient in the controller will be quickly switched to the pre-loaded standby adaptive droop coefficient. S8. The unbalanced power distribution is adjusted by using a PI controller combined with the remaining capacity closed-loop feedback correction circuit, so that the converter station with a higher comprehensive distribution weight can bear more unbalanced power.

2. The active power control method based on adaptive droop control according to claim 1, characterized in that, In step S1, the AC side electrical parameters and DC node voltage are collected in real time through high-precision voltage transformers and current transformers.

3. The active power control method based on adaptive droop control according to claim 1, characterized in that, In step S2, the calculation cycle of the remaining capacity is kept consistent with the collection cycle of the operating parameters to ensure that the remaining capacity data is updated synchronously in real time.

4. The active power control method based on adaptive droop control according to claim 1, characterized in that, The formula for calculating the remaining capacity in step S2 is as follows: ； in, Let be the remaining capacity of the j-th converter station. Let j be the rated power of the j-th converter station. Let the current load power of the j-th converter station be denoted as ; the remaining capacity is processed using a saturation limiting function, and a lower limit protection value for the remaining capacity is set. When the remaining capacity is lower than Time uniform This avoids a sudden change in the droop coefficient when the remaining capacity approaches 0.

5. The active power control method based on adaptive droop control according to claim 1, characterized in that, The formula for calculating the comprehensive allocation weight in step S3 is as follows: ； In the formula This represents the comprehensive allocation weight of the j-th converter station at sampling time k. This represents the remaining capacity after normalized clipping. This represents the maximum rated capacity among all converter stations. The dynamic weights for strong and weak AC systems are represented by a, b, and c, which are positive sensitivity indices corresponding to the adjustment weights of remaining capacity, rated capacity, and AC support capacity, respectively.

6. The active power control method based on adaptive droop control according to claim 1, characterized in that, The dynamic weights of the strong and weak communication system The method of obtaining it is: The AC system support capacity index is calculated by collecting the AC side short-circuit ratio, bus voltage deviation, and current margin of the converter station, and then normalizing and limiting the support capacity index to obtain... The higher the short-circuit ratio of the AC system, the larger the corresponding dynamic weight value.

7. The active power control method based on adaptive droop control according to claim 1, characterized in that, The formula for calculating the standby adaptive droop coefficient in step S4 is as follows: ； In the formula, This represents the standby adaptive droop coefficient calculated at sampling time k for the j-th converter station; Indicates the base sag coefficient; and These are the dynamic lower limit and dynamic upper limit of the droop coefficient for the j-th converter station, respectively; To prevent zero constant.

8. The active power control method based on adaptive droop control according to claim 1, characterized in that, In step S6, the fault prediction indicators include AC voltage change rate, DC voltage fluctuation amplitude, and converter station output current change amplitude. If any one of the indicators exceeds the preset threshold, it is determined that the warning threshold has been reached, and the standby droop coefficient preloading operation is executed.

9. The active power control method based on adaptive droop control according to claim 1, characterized in that, The disturbance events in step S7 include converter station short-circuit faults, planned / unplanned shutdowns of converter stations, DC line open circuits, and sudden increases in AC side load. The overall switching response time of the droop coefficient is no more than 10ms.

10. The active power control method based on adaptive droop control according to claim 1, characterized in that, The remaining capacity closed-loop feedback correction stage in step S8 is implemented as follows: The system collects the actual unbalanced power shared by each converter station in real time, compares it with the target power shared based on the comprehensive allocation weight, and adds the power difference as a correction amount to the standby adaptive droop coefficient to eliminate the allocation deviation caused by long-term load drift.

11. The active power control method based on adaptive droop control according to claim 1, characterized in that, The PI controller proportional gain The integral gain has a value range of 0.5 to 2.

0. The value range is 0.1~0.5; the output of the PI controller, together with the remaining capacity correction, adjusts the active power output of each converter station.

12. The active power control method based on adaptive droop control according to claim 1, characterized in that, The multi-terminal flexible DC transmission system is a four-terminal VSC-HVDC topology, which includes at least two strong AC system-side converter stations and at least two weak AC system-side converter stations; the short-circuit ratio of the strong AC system is greater than a preset threshold, and the short-circuit ratio of the weak AC system is lower than a preset threshold.

13. The active power control method based on adaptive droop control according to claim 1, characterized in that, The sensitivity indices a, b, and c all range from 0.2 to 3.

0. When it is necessary to strengthen the constraint effect of remaining capacity on power allocation, the value of index a is increased; when it is necessary to weaken the influence of AC system support differences, the value of index c is decreased.

Images