Network-configuration type energy storage preventive voltage control method based on voltage sensitivity

Abstract

This invention discloses a preventative voltage control method for grid-connected energy storage based on voltage sensitivity, comprising the following steps: updating the voltage sensitivity matrix online based on real-time operating data of the distribution network and predicting future photovoltaic output curves; determining whether there is a voltage limit exceeding situation at future nodes based on the voltage sensitivity matrix; if there is a voltage limit exceeding situation at future nodes, calculating the active power required by the grid-connected inverter based on the voltage sensitivity matrix and photovoltaic output, and generating a charging curve; adaptively adjusting the charging curve based on the real-time state of charge of the energy storage, and charging the grid-connected inverter before the voltage limit exceeds the limit. The beneficial effects of this invention are: changing the voltage support mode of the energy storage module in the grid-connected energy storage system from passive response to active prevention; when the system's reactive power capacity is insufficient, pre-charging the energy storage allows for sufficient safety margin for the system voltage during critical future periods and maximizes the absorption of photovoltaic power, reducing curtailment.

Description

Technical Field

[0001] This invention relates to the technical field of voltage control, and specifically to a preventive voltage control method for grid-type energy storage based on voltage sensitivity. Background Technology

[0002] With the increasing use of renewable energy in power systems, large-scale integration of distributed photovoltaic (PV) power has become a typical feature of modern distribution networks. However, the intermittency, uncertainty, and volatility of PV power generation pose serious challenges to the safe operation of distribution networks. Especially during periods of high PV power generation, insufficient load absorption capacity can lead to a large amount of PV power being transmitted in reverse, which can easily cause voltage over-limit problems and seriously threaten the reliability of power supply.

[0003] To address this challenge, existing research mainly focuses on reactor compensation, inverter reactive power regulation, and energy storage devices. However, these methods are mostly passive response control. Reactor and inverter compensation suffer from response delays or capacity limitations, while traditional energy storage control and photovoltaic active power reduction typically activate only after the voltage has exceeded its limit. This "limit exceeded first, then correction" approach not only risks curtailment of solar power but also accelerates equipment aging due to frequent switching of energy storage.

[0004] In recent years, grid-forming energy storage (GFES) has provided new ideas for voltage control due to its fast power response and autonomous grid support capabilities. However, existing research on grid-forming energy storage either focuses on its parameter configuration and optimized location or still uses it as a power support after exceeding limits, failing to fully realize its potential for forward-looking and proactive regulation. Summary of the Invention

[0005] Purpose of the invention: To provide a preventive voltage control method for grid-type energy storage based on voltage sensitivity, so as to solve the above-mentioned problems existing in the prior art.

[0006] Technical Solution: A preventive voltage control method for grid-connected energy storage based on voltage sensitivity includes the following steps: updating the voltage sensitivity matrix online based on real-time operating data of the distribution network and predicting future photovoltaic output curves; determining whether there is a voltage limit exceeding situation at future nodes based on the voltage sensitivity matrix; if there is a voltage limit exceeding situation at future nodes, calculating the active power required by the grid-connected inverter based on the voltage sensitivity matrix and photovoltaic output, and generating a charging curve; adaptively adjusting the charging curve based on the real-time state of charge of the energy storage, and performing charging of the grid-connected inverter before the voltage limit exceeds, reserving sufficient voltage safety margin for the system and avoiding the occurrence of limit exceeding.

[0007] Preferably, the voltage sensitivity matrix is ​​obtained by inverting the Jacobian matrix used in the power flow calculation of the Newton-Raphson algorithm, and the voltage sensitivity matrix is: , in, The active power of the node. This represents the change in reactive power. This represents the change in the phase angle of the node voltage. S represents the change in the magnitude of the node voltage. θP S θQ S UP S UQ These are the four elements in the voltage sensitivity matrix.

[0008] Preferably, determining whether a node voltage exceeds its limit based on the voltage sensitivity matrix specifically includes: If U P0 +ΔU nmax ≤1.05pu, the reactive power capacity of grid-type energy storage is sufficient to cope with photovoltaic output, and no preventive voltage control is required; if U P0 +ΔU nmax >1.05 pu, the reactive power capacity of the grid-type energy storage is insufficient, and there is a risk of voltage exceeding the limit. Preventive voltage control is activated; among which U P0 Let ΔU be the voltage at the end node of the feeder when the initial photovoltaic output is P0. nmax This represents the maximum voltage change at node n. The expression for the maximum change in voltage at node n is: , Where, ΔQ GFESmax ΔP represents the maximum reactive power change in grid-type energy storage. PVmax This represents the change in active power output at maximum photovoltaic capacity. , This is the element in the nth row and nth column of the voltage sensitivity matrix.

[0009] Preferably, it also includes a prediction correction mechanism, which re-predicts the photovoltaic output curve in each control cycle, generates a new charging curve reference value, compares the charging curve reference values ​​generated by two adjacent predictions, and determines whether the total charging energy change is less than a threshold ΔE. th The changes in the starting time t0 and ending time t3 of the charging curve are less than the time threshold Δt. th The peak power change of the charging curve is less than the threshold ΔP th At that time, confirm the photovoltaic output curve prediction document and stop updating the charging curve.

[0010] Preferably, when the grid-connected inverter is charging, the voltage cannot fall below the lower voltage limit during the voltage drop phase, and the voltage drop constraint specifically includes: , Where, ΔP GFESrefdemax This is the maximum net active power of the grid-connected inverter during the charging power increase phase.

[0011] Preferably, the adaptive adjustment uses the ampere-hour integral method to calculate the state of charge, divides the voltage safety margin equally, and matches the optimal charging curve corresponding to the maximum rechargeable energy of the energy storage; if the maximum rechargeable energy of the energy storage is less than the minimum equal-divided energy, an alarm is triggered.

[0012] Preferably, the maximum rechargeable energy of the energy storage is: , Among them, E totalmax For the maximum rechargeable energy stored, ΔSOC max E represents the maximum change in SOC of energy storage. btotal Where is the rated capacity of the battery, and k is the overall efficiency coefficient.

[0013] The beneficial effects of this invention are as follows: This invention changes the voltage support mode of energy storage modules in grid-type energy storage systems from passive response to active prevention. When the system's reactive power capacity is insufficient, pre-charging of energy storage ensures sufficient safety margin for system voltage during critical future periods and maximizes photovoltaic absorption, reducing curtailment. Furthermore, by designing a predictive correction mechanism in the control method, this invention continuously updates the charging curve reference value based on predicted photovoltaic output, avoiding erroneous charging of energy storage and making the control strategy robust. Finally, the adaptive control method for energy storage charging power in this invention selects the optimal charging curve based on the energy storage's SOC, maximizing voltage limit prevention while ensuring safe operation of energy storage, thus increasing the economy and sustainability of energy storage. Attached Figure Description

[0014] Figure 1 This is the control flowchart of the present invention; Figure 2 This is a typical daily photovoltaic power output curve for a certain day; Figure 3 A schematic diagram illustrating the charging time and charging amount. Figure 4 A reference value diagram for the charging variation curve of grid-type energy storage; Figure 5 This is a comparison chart of the power reference value and the actual power output curve; Figure 6 Schematic diagram of a preventative voltage control strategy; Figure 7 Flowchart of the control strategy prediction and correction mechanism; Figure 8 Timeline diagram of preventative voltage control strategy; Figure 9 Flowchart of the preventative voltage control strategy; Figure 10 Voltage graphs under different charging curves; Figure 11 Frequency plots under different charging curves; Figure 12 The actual output curves of the active power reference curve under different inertia; Figure 13 A comparison of voltage waveforms under different inertia for the control strategy; Figure 14 This is a graph showing the actual and predicted photovoltaic output curves for Event 1. Figure 15 The voltage curves of the control strategy under actual and predicted photovoltaic conditions are shown. Figure 16 This is a graph showing the actual and predicted photovoltaic output curves for Event 2. Figure 17 The voltage curves are for each voltage without using the predictive correction mechanism; Figure 18 For each voltage curve using the predictive correction mechanism; Figure 19 This is a graph showing the actual and predicted photovoltaic output curves for Event 3. Figure 20 Voltage curves with and without predictive correction mechanism; Figure 21 A comparison chart of SOC with and without adaptive control; Figure 22 A comparison chart showing the use and non-use of adaptive control voltage; Figure 22 The graphs show the SOC variation curves under different adaptive control conditions. Figure 23 The graph shows the adaptive control voltage variation under different conditions. Detailed Implementation

[0015] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Specific Implementation

[0016] As attached Figure 1-3As shown in this embodiment, the preventive voltage control method for grid-type energy storage based on voltage sensitivity includes the following steps: The voltage sensitivity matrix is ​​updated online based on real-time operation data of the distribution network, and the future photovoltaic output curve is predicted. Based on the voltage sensitivity matrix, determine whether there is a voltage limit violation in the future node voltage; If there is a voltage limit violation at a future node, the active power required by the grid-connected inverter is calculated based on the voltage sensitivity matrix and photovoltaic output, and a charging curve is generated. The charging curve is adaptively adjusted based on the real-time state of charge of the energy storage. Before the voltage exceeds the limit, the grid inverter performs charging to reserve sufficient voltage safety margin for the system and avoid the occurrence of exceeding the limit.

[0017] Voltage sensitivity calculation: Given parameters such as distributed photovoltaic power generation, node loads, and line impedance at a certain moment, the voltage sensitivity matrix can be obtained from the Jacobian matrix used in power flow calculations using the Newton-Raphson algorithm: , Inverting the above equation yields the voltage sensitivity matrix: , From the voltage sensitivity matrix, we can obtain: , in, The active power of the node. This represents the change in reactive power. This represents the change in the phase angle of the node voltage. S represents the change in the magnitude of the node voltage. θP S θQ S UP S UQ These are the four elements in the voltage sensitivity matrix, where n is the number of nodes in the distribution network. ΔU i This represents the change in voltage amplitude at the node. and It is the element in the i-th row and j-th column of the voltage sensitivity matrix, ΔP j and ΔQ j Let be the changes in active and reactive power at node j, respectively. If the changes in active and reactive power at all nodes except node n are zero, then the change in node voltage amplitude can be obtained from the above formula as follows: , However, the sensitivity calculation based on the inversion of the Jacobian matrix is ​​essentially a static and linear approximation. In distribution networks with high proportions of photovoltaic power and variable power flow directions, the prediction accuracy of a fixed sensitivity matrix is ​​difficult to guarantee. To address this issue, this invention introduces an "event-triggered sensitivity matrix final update mechanism" in the control method design. By continuously updating the sensitivity matrix online, a final update is performed when a voltage over-limit event is triggered. This ensures the effectiveness of the entire preventative control strategy in dynamic scenarios.

[0018] Grid-based energy storage systems and energy storage charge and discharge control: Currently, grid-connected inverter control methods are mainly divided into two categories: grid-following control and grid-connected control. Unlike grid-following control, grid-connected control does not require a phase-locked loop (PLL) to obtain the grid phase; instead, it controls frequency and voltage based on the differential power flow. Depending on the control components, grid-connected control is further divided into droop control and virtual synchronous generator control, among others. Grid-connected energy storage systems consist of a grid-connected control inverter combined with energy storage batteries.

[0019] The energy storage battery is connected to the DC side of the inverter via a bidirectional DC / DC circuit. The inverter side uses grid-based control. When the grid-based inverter releases power, the battery discharges, and the State of Charge (SOC) decreases; conversely, the battery charges, and the SOC increases. The grid-based inverter can control the release or absorption of power by adjusting its active power reference value Pref (assuming the output is positive), thereby controlling the charging and discharging behavior of the energy storage battery, as shown in Table 1.

[0020] Table 1 Energy Storage Charging and Discharging Behavior Control Table 0 No action No action Almost unchanged + Release power Discharge decline - Absorbed power Charge rise Preventative voltage control methods: Voltage over-limit risk assessment: The typical daily photovoltaic power output curve under ideal sunny weather conditions on a certain day is selected as the research object, such as... Figure 2 As shown.

[0021] Depend on Figure 2 It can be seen that the initial output of the photovoltaic system is P0, the output begins to increase at time t1, and reaches its peak value P at time t2. PVmax Then, at time t3, the output drops to the initial value P0, during which the maximum change in photovoltaic output is ΔP. PVmax Assume that both the photovoltaic (PV) and grid-connected energy storage are connected at node n at the end of the feeder. The load change at node n is 0. When the initial output of the PV is P0, the node voltage is U. P0 From the formula for the change in node voltage magnitude, we can obtain: , Where, ΔP PV and ΔQ PVThese are the changes in active and reactive power of photovoltaic power, ΔP GFES and ΔQ GFES These represent the changes in active and reactive power for grid-based energy storage. Assuming the photovoltaic system operates at a unit power factor and the grid-based energy storage initially has no activity, the above equation can be transformed into: , From the above formula, we can obtain: , Where, ΔU nmax Let ΔQ be the maximum voltage change at node n. GFESmax ΔP represents the maximum reactive power change in grid-type energy storage (positive indicates the release of inductive reactive power, negative indicates the absorption of inductive reactive power). PVmax This represents the change in active power output at maximum photovoltaic capacity. , This is the element in the nth row and nth column of the voltage sensitivity matrix.

[0022] Determining whether a node voltage exceeds its limit based on the voltage sensitivity matrix specifically includes: If U P0 +ΔU nmax If the reactive power of the grid-type energy storage is ≤1.05 pu, then the reactive power capacity is sufficient to cope with the photovoltaic output, and no preventative voltage control is required. If U P0 +ΔU nmax If the voltage exceeds 1.05 pu, the reactive power capacity of the grid-type energy storage will be insufficient, and the voltage will exceed the limit, requiring the activation of preventive voltage control.

[0023] Preventative voltage control methods: The essence of preventative voltage control methods: It is known that the reactive power-voltage control loop of a grid-type converter typically employs a "reactive power-voltage droop control" method, i.e.: , Among them, Q GFES Q0 represents the reactive power output of grid-connected energy storage, and K represents the initial reactive power output of grid-connected energy storage. q U is the droop factor, U0 is the voltage reference value, typically 1.0 pu, U n This represents the actual value of the current voltage. The essential formula of the above preventative voltage control method can be transformed to obtain: , WhenU P0 +ΔU nmax When the value is greater than 1.05 pu, the change in the n-node reference voltage ΔU is introduced. nref To satisfy: , Where ΔU bThis represents the maximum voltage change at the node after preventative control. From this formula, the range of ΔUnref can be obtained as follows: , From the above formula, we can obtain: , Where ΔP PVref For ΔU nref The corresponding change in photovoltaic power output. The preventative voltage control method essentially shifts P0 upwards to the target power P. PVref At this point, P PVref The voltage amplitude at point P1 is replaced with the voltage amplitude at point P0 to deal with voltage exceeding the limit.

[0024] Design of the energy storage charging power reference curve: This invention designs the charging power reference value of grid-based energy storage as a linear curve, mainly based on the following considerations: 1. Grid friendliness: Linear charging and discharging can ensure smooth and gradual power change, avoiding power jumps caused by constant power switching.

[0025] 2. Equipment protection: Power step changes can cause drastic fluctuations in battery current, accelerating aging; at the same time, it can also easily lead to saturation of the converter current loop.

[0026] 3. Engineering advantages: The linear curve is defined by only three parameters: initial value, terminal value and climbing time. The model is simple, the computational burden is small, and it is easy to generate online and track accurately.

[0027] Design principles of preventative voltage control methods: From the above formula P PVref By adjusting the photovoltaic output curve, P0 can be shifted upwards to the target power P. PVref Charging time and charging amount at the time of application, such as Figure 3 As shown.

[0028] Depend on Figure 3 It can be seen that: Let the photovoltaic output reach P PVref When, the corresponding time is t m and t n When S1 = S2, then from the diagram we can obtain: , We can obtain the following from the above formula: , Where t0 is the charging start time, P PV(t) This is the functional expression for the photovoltaic power output curve.

[0029] The reference value for the charging variation curve of grid-type energy storage is as follows: Figure 4 As shown.

[0030] Figure 4 In this context, Etotal represents the reference value for the total energy absorbed by the inverter, and is determined by... Figure 4 The active power reference value P of the grid inverter can be obtained. GFESref The curve of change, since the output is assumed to be positive and the input to be negative, therefore needs to be adjusted. Figure 4 The negative value of the curve represents the active power reference value P of the actual grid-connected inverter. GFESref The curve shows the change in active power. However, in actual system operation, due to factors such as the virtual inertia of the grid-connected inverter, the actual power output curve of the inverter will not be exactly the same as the active power reference value P. GFESref The change curve is as follows Figure 5 As shown.

[0031] Depend on Figure 5 It can be seen that the actual output power of grid-based energy storage deviates slightly from the reference value at specific times, but the overall trend is consistent, especially when the photovoltaic output is at its maximum, the charging power can still remain stable at P. pvref The power curve remains effective in supporting the proposed voltage control method. Subsequent simulations also show that although factors such as inertia introduce waveform differences, they do not affect the overall effectiveness of the preventative control method.

[0032] The principle of preventive voltage control strategy is based on Figure 6 It can be seen that without preventative control methods, the voltage change when the photovoltaic system reaches its maximum output is U. P0 +ΔU a +ΔU b The preventative control strategy shifts P0 upwards to P... PVref At this point, the voltage change is U when the photovoltaic power output is at its maximum. P0 +ΔU b .

[0033] Control strategy prediction and correction mechanism: Since there may be a mismatch between the predicted photovoltaic output and the actual photovoltaic output in practice, if the charging curve is still set according to the original predicted photovoltaic curve, it may cause the control method to fail. To solve this problem, this invention designs a control method prediction correction mechanism. In each control cycle, the photovoltaic output curve is re-predicted, a new charging curve reference value is generated, and the charging curve reference values ​​generated by two adjacent predictions are compared. If the following conditions are met: When the change in total charging energy is less than the threshold ΔE th、 The changes in the charging curve at the start time t0 and the end time t3 are less than the time threshold Δt. th The peak power change of the charging curve is less than the threshold ΔP th At that time, confirm the photovoltaic output curve prediction document and stop updating the charging curve.

[0034] When the predicted charging curve stabilizes, the inverter will perform a final prediction check just before executing the final charging curve. If the conditions are still met, the prediction will proceed; otherwise, the update will continue. The control strategy prediction and correction mechanism flow is as follows: Figure 7 As shown.

[0035] Preventive voltage control method flow: From the above, the entire control method flow can be obtained as follows: At the current time t c Based on the updated voltage sensitivity matrix and photovoltaic output prediction, it determines whether the future voltage will exceed the limit. If it does, the matrix and prediction data are immediately updated to generate a charging curve, and the grid-connected inverter prepares to execute. Just before execution, the voltage sensitivity matrix and charging curve are updated for the final command. Subsequently, the inverter begins charging at time t0, with power increasing linearly from 0 and voltage decreasing accordingly. Around time t1, as photovoltaic output begins, the voltage gradually recovers. Around time t... m At that moment, the charging power reaches its maximum P. PVref The voltage rises accordingly to U P0 In t m To t n During the period, the charging power is at P PVref The voltage fluctuated slightly in the vicinity, during which time it ranged from U P0 Change. t n Around that time, as photovoltaic output decreases, the charging power decreases from P PVref The power output begins to decrease, and at time t3, the photovoltaic system stops outputting power, and the energy storage charging ends. See the complete timeline for details. Figure 8 .

[0036] When the grid-connected inverter is charging, the voltage cannot fall below the lower voltage limit during the voltage drop phase. The voltage drop constraint specifically includes: , Where, ΔP GFESrefdemax This is the maximum net active power of the grid-connected inverter during the charging power increase phase.

[0037] Maximum ramp rate and maximum power constraints for energy storage: , Among them, P batcr max P batdr max and P bat max These are the maximum ramp rate of energy storage, the maximum ramp rate of energy storage, and the maximum charging power of energy storage, respectively.

[0038] Quantitative indicators of control performance: To quantify the superiority of the control method presented in this paper in reducing the curtailment rate and the number of energy storage operations, the curtailment rate calculation formula is defined as follows: , Among them, S pvcr For the light rejection rate, t f0 and t f1 These are the start and end times of the recording, P. PVtheomax(t) P represents the theoretical maximum output of the photovoltaic system at time t. PVactual (t) represents the actual photovoltaic output at time t. The number of energy storage actions is defined as the instantaneous moment when the power state of the energy storage device switches, counted as one action. That is, from the idle state (P... ess =0) Switch to charging state (P ess A change of <0 is recorded as one action, and switching from charging state to idle state is also recorded as one action. In actual measurement, to avoid false triggering due to measurement noise, it is set that only when the absolute value of power changes by more than a certain value is it considered a valid state switch, and this is recorded as one energy storage action.

[0039] Preferably, the adaptive adjustment uses the ampere-hour integral method to calculate the state of charge, divides the voltage safety margin equally, and matches the optimal charging curve corresponding to the maximum rechargeable energy of the energy storage; if the maximum rechargeable energy of the energy storage is less than the minimum equal-divided energy, an alarm is triggered.

[0040] The adaptive control method for energy storage charging power specifically includes: Through ΔU nmax To set ΔU nref , through ΔU nref Determine P PVref This determines that P0 is translated upwards to P PVref Reference values ​​for the charging curve and total charging energy E at the location total However, if a fixed ΔU is used nref The current configuration can easily lead to overcharging of energy storage. To address this issue, this invention proposes an adaptive control method for energy storage charging power. This method automatically matches the optimal charging curve based on the SOC state of the energy storage, thereby avoiding overcharging.

[0041] The ampere-hour integration method is a commonly used method for calculating the state of charge (SOC) of a battery. The basic principle of this method is to calculate the SOC by accumulating the amount of electricity charged and discharged during the charging and discharging of the battery.

[0042] By the ampere-hour integration method, we can obtain: , Among them, C n The rated capacity of the battery (unit: Ah), t start and t end These represent the start and end times of charging, respectively, where η is the coulombic efficiency, and i bat (t) represents the battery terminal current. Transforming the ampere-hour integration method, we get: , Where k is the overall efficiency coefficient, U bat E is the battery terminal voltage. total Total charging energy (unit: kWh), E btotal Rated capacity of the battery (unit: kWh).

[0043] From the above, we can obtain ΔU nrefmin =ΔU nmax +U P0 -1.05pu, and ΔU b The selection range is 0 ≤ ΔU b ≤1.05pu-U P0 Therefore, ΔU bmax =1.05pu-U P0 The adaptive principle is as follows: ΔU bmax The system is divided into N equal parts, and the total charging energy for each equal part is calculated. The maximum rechargeable energy is obtained from the energy storage SOC, and then the optimal charging curve is matched. The design is as follows: , When ΔU is obtained nref When the voltage is equal to different equal parts, the corresponding total charging energy is E respectively. bat1 E bat2 E bat3 ...E batN Based on the above, the maximum charging energy of the energy storage is: , Among them, E totalmax For the maximum rechargeable energy stored, ΔSOC max This represents the maximum change in SOC of the energy storage. Then, E is determined. totalmax Located in E bat1 E bat2 E bat3 ...E batN Within which interval, if E bath ≤E totalmax ≤E bat(h+1) Then E totalmax =E bath The inverter then executes E bath The corresponding charging curve, if E totalmax <E bat1 This indicates that the energy storage is still exceeding the charging voltage limit even at its maximum capacity, thus triggering an alarm. Due to factors such as virtual inertia in grid-connected inverters, the actual energy E absorbed by the inverter during the entire charging process is limited. actual Often less than E total Therefore, E is used. total When designing for maximum charging energy, if Etotal It will not lead to overcharging of energy storage, so the actual E actual This further reduces the risk of overcharging of energy storage. As N increases, the SOC can be fully utilized, but the operating speed will decrease.

[0044] Based on the above, the flow chart for the preventive voltage control method is as follows: Figure 1 As shown. Specific Implementation Example 2 To verify the effectiveness of the proposed control strategy, a simulation model was built on Matlab / Simulink based on the proposed strategy. For the grid-type control, virtual synchronous generator control was selected, and the parameters used in the model are shown in Table 2.

[0046] Table 2 System Parameters Initial voltage value <![CDATA[U P0 / p.u.]]> 1.0 Inverter initial active power reference value <![CDATA[P GFESref / W]]> 0 Rated capacity of energy storage <![CDATA[E btotal / Kwh]]> 0.4 Photovoltaic maximum output <![CDATA[P PVmax / Kw]]> 200 Maximum voltage change <![CDATA[ΔU nmax / p.u.]]> 0.065 Initial state of charge of energy storage SOC / % 50.5 In the model, the photovoltaic output starts to increase from 1s, reaches the maximum output of 200kW at 1.5s, then decreases and recovers to the initial output at 2s. ΔUnref = 0.054pu is set.

[0047] Linear charging curve verification: To verify the rationality of the linear charging curve proposed in this invention, the same maximum charging power was set, and the linear charging curve was compared with the constant power charging curve, such as... Figure 9 , 10 As shown, Figure 9 , 10 A comparison of voltage and frequency under different charging curves.

[0048] Under the same maximum charging power, by Figure 9 It can be seen that constant power charging experiences sudden power changes at the start and stop, resulting in significant voltage drops and increases, respectively; while linear charging maintains a stable voltage throughout the entire process without significant sudden changes. Meanwhile, due to... Figure 10 It is known that sudden power changes during constant power charging can cause sudden changes in system frequency, while the frequency changes during linear charging are more stable. Therefore, the linear charging curve selected in this invention is superior to the constant power charging curve in terms of both voltage control and system frequency, and is more conducive to maintaining the stability of the power system during charging.

[0049] Verification of the influence of different inertia on voltage control: To verify the impact of the control strategy in this paper on voltage control under different inertia conditions, the virtual inertia J and damping D corresponding to inertia 1, 2, 3, 4, and 5 are set as shown in Table 3 below.

[0050] Table 3 Different inertia values Inertia 1 0.05 10 Inertia 2 0.5 60 Inertia 3 2 100 Inertia 4 5 150 Inertia 5 10 250 The actual output power curves of the inverter under different inertia and the corresponding voltage control curves under the same inertia are as follows: Figure 11 , 12 As shown, Figure 11 , 12 This diagram illustrates the impact of different inertia rates on control.

[0051] Depend on Figure 11 It can be seen that, due to the inertia of the grid-connected inverter, its actual output power curve differs from the given active power reference value P. GFESref The curves often cannot perfectly coincide, and the actual output curves differ depending on the inertia of the grid inverter. Meanwhile, due to... Figure 12 It can be seen that although the actual charging curve is similar to the given active power reference value P, GFESref Although the curves differ, their overall trend is the same. Aside from slight variations during the initial voltage drop phase of charging, they all reach near maximum active power within the range of maximum photovoltaic output. Therefore, even with virtual inertia in the inverter, etc., Figure 11 and Figure 12 It is also clear that the preventive voltage control strategy proposed in this invention is still effective.

[0052] Prediction correction mechanism and robustness verification To verify the robustness of the control strategy prediction and correction mechanism and the control strategy in this paper, three events were selected for simulation verification, as shown in Table 4 below.

[0053] Table 4 Description of the three types of events 1 The actual photovoltaic power output curve waveform differs greatly from the prediction. 2 There is a difference between the actual and predicted photovoltaic power output time. 3 The actual photovoltaic output differs from the prediction in both output time and waveform. The events and simulation results are as follows: Figures 13-19 As shown.

[0054] Depend on Figure 13 and Figure 14 It can be seen that even when there is a significant difference between the predicted photovoltaic curve and the actual photovoltaic curve, the control strategy of this invention remains effective. This is because the control method of this invention is mainly based on the maximum photovoltaic output. Therefore, when the set charging power is reached at the maximum photovoltaic output, the voltage will not exceed the limit. Thus, as long as the maximum photovoltaic output is known, waveform changes outside of the maximum output have no impact on the control described in this invention. Figure 15 , Figure 16 and Figure 17 It is known that when the actual photovoltaic output time deviates significantly from the predicted time, the control method will fail. This is because without a predictive correction mechanism, the charging curve is fixed by the initially predicted curve. When the maximum charging power is reached, the actual photovoltaic output is not high enough, leading to incorrect charging of the energy storage and causing a significant voltage drop, even below the lower limit. The longer the actual photovoltaic delay, the more severe the problem. However, with a predictive correction mechanism, the charging curve is continuously updated based on the photovoltaic output curve, thus preventing incorrect charging of the energy storage and keeping the voltage within a reasonable range. Figure 18 and Figure 19 It can be seen that when the actual photovoltaic performance differs from the prediction in both curve and time, the control strategy in this paper is still effective, indicating that the control strategy in this paper has a certain robustness in voltage control when there is an error between the prediction and the actual performance.

[0055] Adaptive control strategy verification To verify the effectiveness of the adaptive control strategy, N was set to 5, and the simulation curves are as follows. Figures 20-23 As shown.

[0056] Depend on Figure 20 It can be seen that when the initial SOC is set to 76%, without the adaptive control strategy, the SOC eventually stabilizes at 81.3%, exceeding the maximum safe upper limit threshold of 80% for energy storage. However, with adaptive control, the optimal charging curve ΔU is automatically matched based on 4% of the maximum usable SOC capacity. nref The charging curve shows a rate of 0.035 PU, at which point the SOC finally stabilizes at 79.9%. Meanwhile, from... Figure 21 It can be seen that the voltage safety margin without adaptive control is higher than that with adaptive control. This is because, in order to avoid overcharging of the energy storage, the adaptive control automatically matches the charging curve corresponding to the maximum available capacity of the energy storage. The maximum charging power is lower than the maximum charging power without adaptive control, so the voltage margin is relatively lower.

[0057] Figure 22 and Figure 23 In cases 1, 2, and 3, the SOC is 77.5%, 76%, and 74.5%, respectively. Figure 22 It can be seen that when the initial SOC of energy storage is different, the adaptive control can automatically match the optimal charging curve under that SOC condition, and by Figure 23 Therefore, the voltage safety margin varies depending on the maximum available capacity of the energy storage. The larger the maximum available capacity of the energy storage, the larger the voltage safety margin, and vice versa.

[0058] To address the issue of voltage exceedance caused by photovoltaic (PV) power in grid-based energy storage systems under conditions of insufficient reactive power capacity, this invention proposes a preventative voltage control method for grid-based energy storage systems based on voltage sensitivity. Through theoretical modeling and simulation verification, the following conclusions are drawn: This invention transforms the voltage support mode of energy storage modules in grid-based energy storage systems from passive response to active prevention. When the system's reactive power capacity is insufficient, pre-charging the energy storage ensures sufficient safety margin for system voltage during critical future periods and maximizes PV power absorption, reducing curtailment. Furthermore, by designing a predictive correction mechanism in the control method, this invention continuously updates the charging curve reference value based on predicted PV output, avoiding erroneous charging of the energy storage and making the control strategy robust. Finally, the adaptive charging power control method of this invention selects the optimal charging curve based on the energy storage's state of charge (SOC), maximizing voltage exceedance prevention while ensuring safe operation of the energy storage, thus increasing the economy and sustainability of energy storage.

[0059] The preferred embodiments have been shown and described, but should not be construed as limiting the invention itself. Various changes in form and detail may be made without departing from the spirit and scope of the invention as defined in the appended claims.

Claims

1. A network configuration type energy storage preventive voltage control method based on voltage sensitivity, characterized in that: Includes the following steps: The voltage sensitivity matrix is ​​updated online based on real-time operation data of the distribution network, and the future photovoltaic output curve is predicted. Based on the voltage sensitivity matrix, determine whether there is a voltage limit violation in the future node voltage; If there is a voltage limit violation at a future node, the active power required by the grid-connected inverter is calculated based on the voltage sensitivity matrix and photovoltaic output, and a charging curve is generated. The charging curve is adaptively adjusted based on the real-time state of charge of the energy storage. Before the voltage exceeds the limit, the grid inverter performs charging to reserve sufficient voltage safety margin for the system and avoid the occurrence of exceeding the limit.

2. The voltage-sensitivity-based network-forming energy-storage preventive voltage control method according to claim 1, characterized by: The voltage sensitivity matrix is ​​obtained by inverting the Jacobian matrix used in the power flow calculation of the Newton-Raphson algorithm. The voltage sensitivity matrix is ​​as follows: ， wherein, P is the active power of the node, Q is the reactive power variation, Θ is the node voltage phase angle variation, V is the node voltage magnitude variation, S θP , S θQ , S UP , S UQ are four elements in the voltage sensitivity matrix.

3. The voltage-sensitivity-based network-forming energy-storage preventive voltage control method according to claim 2, characterized by: Determining whether a node voltage exceeds its limit based on the voltage sensitivity matrix specifically includes: If U P0 +ΔU nmax ≤1.05pu, the reactive power capacity of grid-type energy storage is sufficient to cope with photovoltaic output, and no preventive voltage control is required; if U P0 +ΔU nmax >1.05 pu, the reactive power capacity of the grid-type energy storage is insufficient, and there is a risk of voltage exceeding the limit. Preventive voltage control is activated; among which U P0 Let ΔU be the voltage at the end node of the feeder when the initial photovoltaic output is P0. nmax This represents the maximum voltage change at node n. The expression for the maximum change in voltage at node n is: ， Where, ΔQ GFESmax ΔP represents the maximum reactive power change in grid-type energy storage. PVmax This represents the change in active power output at maximum photovoltaic capacity. , This is the element in the nth row and nth column of the voltage sensitivity matrix.

4. The preventive voltage control method for grid-type energy storage based on voltage sensitivity according to claim 3, characterized in that: It also includes a prediction correction mechanism, which re-predicts the photovoltaic output curve in each control cycle, generates a new charging curve reference value, compares the charging curve reference values ​​generated by two adjacent predictions, and determines whether the total charging energy change is less than the threshold ΔE. th The changes in the starting time t0 and ending time t3 of the charging curve are less than the time threshold Δt. th The peak power change of the charging curve is less than the threshold ΔP th At that time, confirm the photovoltaic output curve prediction document and stop updating the charging curve.

5. The voltage-sensitivity-based network-forming energy-storage preventive voltage control method according to claim 4, characterized by: When the grid-connected inverter is charging, the voltage cannot fall below the lower voltage limit during the voltage drop phase. The voltage drop constraint specifically includes: ， where ΔP GFESrefdemax is the maximum net active power of the grid-forming inverter in the charging power ramp-up phase.

6. The voltage-sensitivity-based network-forming energy-storage preventive voltage control method according to claim 5, characterized by: The adaptive adjustment uses the ampere-hour integral method to calculate the state of charge, divides the voltage safety margin equally, and matches the optimal charging curve corresponding to the maximum rechargeable energy of the energy storage; if the maximum rechargeable energy of the energy storage is less than the minimum equal division energy, an alarm is triggered.

7. The voltage-sensitivity-based network-forming energy-storage preventive voltage control method according to claim 6, characterized by: The maximum rechargeable energy of the energy storage is: ， Wherein, E totalmax is the maximum chargeable energy of the energy storage, ΔSOC max is the maximum SOC change amount of the energy storage, E btotal is the rated capacity of the battery, and k is the comprehensive efficiency coefficient.

Images