A constant-power soc equalization method for a modular multilevel battery energy storage system

Abstract

The application discloses a kind of modular multilevel battery energy storage system's fixed power SOC equalization method.The method first obtains the current and initial SOC value of each submodule battery and system total power, and calculates the average SOC of system current and initial;Then respectively calculate the fine tuning power correction value based on fixed coefficient method and the forced equalization power correction value based on fixed power method;Subsequently set the inner and outer boundary threshold of hysteresis control, and combine system equalization mode to carry out logical judgment, between the smooth switching of two correction values to select target power correction value;Finally, the selected target power correction value is added to the reference distribution power to generate target reference power, which is converted into reference current instruction and introduced into the current inner loop control of bidirectional DC / DC converter, to generate duty cycle drive switch tube.The application completely eliminates the traditional three-level hierarchical equalization step, realizes single-stage fast equalization, greatly reduces the chip computing power burden, and effectively avoids the high-frequency buffeting of system at equalization critical point.

Description

Technical Field

[0001] This invention relates to the fields of power electronics technology and distributed energy storage system control, specifically to a constant power SOC equalization method for a modular multilevel battery energy storage system. Background Technology

[0002] With the increasing demand for grid connection of renewable energy, battery energy storage systems based on modular multilevel converters have become a key technology in the field of large-capacity energy storage due to their high voltage level, high power transmission capability, and the flexibility of modular design. However, due to differences in manufacturing processes and initial conditions of individual battery cells, as well as varying degrees of aging during long-term charge-discharge cycles, an imbalance in the state of charge among the battery cells is inevitable. Without proactive equalization intervention, some batteries may experience overcharging or over-discharging, which will not only lead to premature battery failure and shorten the overall lifespan of the energy storage system, but also severely reduce the capacity utilization rate of the entire battery pack.

[0003] In existing MMC-BESS topologies, the industry has begun to adopt an integrated approach where the battery is connected in parallel to the half-bridge submodule capacitor via a non-isolated bidirectional DC / DC converter. This structure achieves electrical decoupling between the battery and the submodule capacitor, effectively blocking the impact of low-frequency harmonic currents on the battery, while avoiding the high cost and additional switching losses associated with using isolated DC / DC converters. However, designing a fast and efficient SOC equalization control strategy based on this hardware topology remains a major challenge in the current technological field. Existing SOC equalization control methods generally suffer from the following technical shortcomings:

[0004] 1. It heavily relies on a three-level hierarchical equalization architecture of "phase-to-bridge arm-to-submodule". The coupling and control delay between the levels result in extremely slow overall system equalization.

[0005] 2. When using model prediction or complex feedforward control, the increasing number of sub-modules can easily lead to an explosion in control dimensions, severely increasing the computing burden on the main control chip and easily interfering with the stability of the main circuit.

[0006] 3. The traditional single-coefficient method has insufficient output power when the initial deviation is large, resulting in a long equalization process. When the deviation is small, forced adjustment is very likely to cause overshoot and high-frequency oscillation of the system. Summary of the Invention

[0007] To address the aforementioned technical deficiencies, this invention provides a constant-power SOC equalization method for a modular multilevel battery energy storage system, comprising the following steps:

[0008] S1) Real-time acquisition of the current state of charge (SOC) value of each sub-module battery in the energy storage system. i The initial state of charge (SOC) at the initial operating moment of the system.i_init And the total rated power P of the entire battery energy storage system B Where i is the sequence number of the corresponding submodule, i = 1, 2, 3, ..., 6N, and N is the number of submodules connected in series in each bridge arm;

[0009] S2) According to the SOC i and SOC i_init Calculate the average SOC value of all batteries in the current system. ave and the initial average SOC value at the initial running time of the system. ave_init ;

[0010] S3) Based on the SOC i SOC ave SOC i_ and SOC ave_init Calculate the fine-tuning power correction value ΔP based on the "fixed coefficient method". i And the forced equalization power correction value ΔP based on the "constant power method". solidi ;

[0011] S4) Set the inner and outer boundary thresholds of the hysteresis control and determine the equilibrium mode state variable z of the system in the current control cycle;

[0012] S5) Based on the state variable z determined in step S4) and the set inner and outer boundary thresholds, perform a logical judgment on the ΔP. i and ΔP solidi Smoothly switch between them, select the target power correction value of the submodule in the current control cycle, and update the state variable z;

[0013] S6) Add the selected target power correction value to the reference allocated power of the submodule to generate the target reference power P of the submodule. b * ;

[0014] S7) Convert the target reference power into a reference current command;

[0015] S8) The reference current command is introduced into the current inner loop control of the non-isolated bidirectional DC / DC converter, thereby generating a PWM duty cycle D to drive the switching transistor.

[0016] In step S3), the fine-tuning power correction value based on the constant coefficient method is: ΔP i = k i × (SOC i -SOC ave The forced equalization power correction value based on the constant power method is: ΔP solidi = ki × (SOC i_init -SOC ave_init ); where k i This is the preset balance ratio coefficient.

[0017] In step S4), the specific steps for setting the inner and outer boundary thresholds of the hysteresis control and determining the equilibrium mode state variable z of the system in the current control cycle are as follows:

[0018] Step S4-1) Set the internal narrow threshold of hysteresis control to b and the external wide threshold to kb; where b is the preset minimum allowable steady-state deviation and k is a shake-proof margin coefficient greater than 1.

[0019] Step S4-2) During the initial system runtime, calculate the initial absolute SOC deviation of the submodule |SOC i_init -SOC ave_init |; If the initial deviation is greater than or equal to the external wide threshold kb, then assign an initial value z = ΔP to the state variable. solidi This means determining that the system has initially entered a constant power equilibrium mode; if the initial deviation is less than kb, then the initial value of the state variable z = ΔP is assigned. i That is, to determine that the system initially enters a constant coefficient equilibrium mode;

[0020] Step S4-3) In subsequent control cycles after the system is running normally, directly read the value of the state variable z saved at the end of the previous control cycle; if z = ΔP solidi If z = ΔP, then the system used a constant power equalization method in the previous control cycle; i If so, it is determined that the system used a fixed-coefficient equilibrium method in the previous control cycle;

[0021] In step S5), the specific steps for making logical judgments based on the state variable z determined in step S4) and the set inner and outer boundary thresholds are as follows:

[0022] Step S5-1) Calculate the current SOC absolute deviation of each submodule. diffi = |SOC i - SOC ave |;

[0023] Step S5-2) If the state variable z = ΔP determined in step S4) solidi If the system used constant power equalization in the previous control cycle, then the current absolute SOC deviation (SOC) is determined. diffi Is it less than the internal narrow threshold b? If so, then switch the output fine-tuning power correction value ΔP in the current cycle. i And update the state variable z to ΔP iIf not, maintain the output forced equalization power correction value ΔP. solidi ;

[0024] Step S5-3) If the state variable z = ΔP determined in step S4) i If the system used a fixed-coefficient balancing method in the previous control cycle, then the current absolute SOC deviation (SOC) is determined. diffi Is it less than the external wide threshold kb? If so, maintain the output fine-tuning power correction value ΔP in the current cycle. i If not, switch the output forced equalization power correction value ΔP. solidi And update the state variable z to ΔP. solidi ;

[0025] In step S6), the target reference power P of the submodule is generated. b * The specific steps are as follows:

[0026] Step S6-1) The total rated power P B The power is evenly distributed among the 6N sub-modules of the system to obtain the baseline power distribution P. B / (6N);

[0027] Step S6-2) Add the reference allocated power to the target power correction value selected in step S5) to obtain the target reference power P. b * .

[0028] In step S7), the specific method for converting the target reference power into a reference current command is as follows: the target reference power P b * Divide by the real-time voltage u of the current submodule battery b The reference current command i is obtained. b * .

[0029] In step S8), the specific steps for introducing the reference current command into the inner current loop control of the non-isolated bidirectional DC / DC converter, thereby generating the drive signal for the switching transistor of the non-isolated bidirectional DC / DC converter, are as follows:

[0030] Step S8-1) Set the reference current command i b * Compared with the actual sampled battery current i b The error between them is fed into the current inner loop proportional-integral controller of the non-isolated bidirectional DC / DC converter to calculate the output PWM duty cycle D = (i b *- i b (K) p +Ki / s), where K p and K i These are the proportional coefficient and integral coefficient in the proportional-integral controller, respectively, and s is the Laplace operator;

[0031] Step S8-2) Based on the duty cycle D, a drive signal is generated using complementary pulse width modulation to control the on and off states of switches T3 and T4, thereby controlling the actual battery current i. b Track the reference current command i b * This enables rapid balancing of single-level SOC.

[0032] The beneficial effects of this invention are as follows:

[0033] 1. By directly performing submodule-level power control through a non-isolated bidirectional DC / DC converter, the traditional phase-to-phase and bridge arm equalization steps are completely eliminated, greatly simplifying the control architecture and significantly improving the equalization speed.

[0034] 2. By adopting decentralized underlying distributed logic, the target power is directly converted into current commands and introduced into the inner loop control. The algorithm complexity does not explode with the scale of the system, which greatly reduces the chip's computing power burden and ensures stable operation.

[0035] 3. The introduction of a "constant power + constant coefficient" hybrid mode and internal and external dual threshold hysteresis logic not only overcomes the shortcomings of the single constant coefficient method in the later stage of adjustment and shortens the equalization time, but also effectively avoids the high-frequency state chattering of the system at the equalization critical point. Attached Figure Description

[0036] Figure 1 This is a schematic diagram illustrating the principle of the present invention;

[0037] Figure 2 This is a schematic diagram of the main circuit topology of the modular multilevel converter battery energy storage system used in the embodiments of the present invention.

[0038] Figure 3 The flowchart of hybrid power control based on hysteresis logic proposed in the embodiments of the present invention is shown below;

[0039] Figure 4 This is a block diagram of the battery SOC dual-closed-loop equalization control structure of the non-isolated bidirectional DC / DC converter in an embodiment of the present invention;

[0040] Figure 5 This is a comparison diagram of the power correction principle under different control methods in the embodiments of the present invention;

[0041] Figure 6 The waveform diagram shows a comparison of the single-phase SOC equalization simulation effects of the embodiments of the present invention and the traditional control method;

[0042] Figure 7 The waveform diagram shows a comparison of the single-phase SOC equalization experimental effects of the embodiments of the present invention and the traditional control method. Detailed Implementation

[0043] 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.

[0044] Depend on Figure 2 As shown, the modular multilevel converter battery energy storage system upon which this invention relies adopts a symmetrical three-phase dual-bridge structure. Each bridge arm consists of N energy storage sub-modules connected in series with the bridge arm inductor L, where N is the number of sub-modules connected in series in each bridge arm. Unlike traditional half-bridge sub-modules, the energy storage sub-modules of this invention connect the battery in parallel to the capacitor side of the sub-module through a non-isolated bidirectional DC / DC converter.

[0045] A single energy storage submodule includes four IGBT switching transistors T1, T2, T3, and T4, four anti-parallel diodes, a shared capacitor C, a filter inductor L, and a battery cell. Its internal connections are as follows:

[0046] Switches T1 and T2 are connected in series to form the main circuit port of the half-bridge structure, which is connected to the bridge arm of the MMC.

[0047] A shared capacitor C is connected in parallel across the two ends of the series branch of T1 and T2;

[0048] Switches T3 and T4 are connected in series to form the bridge arm of a non-isolated bidirectional DC / DC converter. The series branch of T3 and T4 is also connected in parallel across the shared capacitor C.

[0049] After the battery cell is connected in series with the filter inductor, one end of it is connected to the midpoint of the series connection of switching transistors T3 and T4, and the other end is connected to the negative terminal of the shared capacitor C.

[0050] With the above structure, the battery is safely decoupled and connected in parallel across the capacitor. T1 and T2 are used to control the output voltage U of the submodule. SM Specifically, when T1 is on and T2 is off, the submodule is engaged, and the output voltage is the capacitor voltage U. C When T1 is off and T2 is on, the submodule is disconnected, and the output voltage is 0. In addition, T3 and T4 act as independent power actuators, which are specifically responsible for adjusting the charging and discharging power of the battery under software closed-loop control, providing a hardware foundation for the subsequent realization of single-stage SOC fast balancing without phase-to-phase and bridge arm cascading steps.

[0051] Based on the above hardware topology, by Figure 1 As shown, this invention provides a constant-power SOC equalization method for a modular multilevel battery energy storage system, characterized by the following steps:

[0052] S1) Real-time acquisition of the current state of charge (SOC) value of each sub-module battery in the energy storage system. i The initial state of charge (SOC) at the initial operating moment of the system. i_init And the total rated power P of the entire battery energy storage system B Where i is the sequence number of the corresponding submodule, i = 1, 2, 3, ..., 6N, and N is the number of submodules connected in series in each bridge arm;

[0053] S2) According to the SOC i and SOC i_init Calculate the average SOC value of all batteries in the current system. ave and the initial average SOC value at the initial running time of the system. ave_init ;

[0054] S3) Based on the SOC i SOC ave SOC i_ and SOC ave_init Calculate the fine-tuning power correction value ΔP based on the "fixed coefficient method". i And the forced equalization power correction value ΔP based on the "constant power method". solidi In step S3), the fine-tuning power correction value based on the constant coefficient method is: ΔP i = k i × (SOC i - SOC ave The forced equalization power correction value based on the constant power method is: ΔP solidi = k i × (SOC i_init - SOC ave_init ); where k i This is the preset equilibrium ratio coefficient;

[0055] S4) Set the inner and outer boundary thresholds of the hysteresis control and determine the equilibrium mode state variable z of the system in the current control cycle; Step S4) specifically includes three steps:

[0056] Step S4-1) Set the internal narrow threshold of hysteresis control to b and the external wide threshold to kb; where b is the preset minimum allowable steady-state deviation and k is a shake-proof margin coefficient greater than 1.

[0057] Step S4-2) During the initial system runtime, calculate the initial absolute SOC deviation of the submodule |SOC i_init -SOC ave_init |; If the initial deviation is greater than or equal to the external wide threshold kb, then assign an initial value z = ΔP to the state variable. solidi This means determining that the system has initially entered a constant power equilibrium mode; if the initial deviation is less than kb, then the initial value of the state variable z = ΔP is assigned. i That is, to determine that the system initially enters a constant coefficient equilibrium mode;

[0058] Step S4-3) In subsequent control cycles after the system is running normally, directly read the value of the state variable z saved at the end of the previous control cycle; if z = ΔP solidi If z = ΔP, then the system used a constant power equalization method in the previous control cycle; i If so, it is determined that the system used a fixed-coefficient equilibrium method in the previous control cycle;

[0059] S5) Based on the state variable z determined in step S4) and the set inner and outer boundary thresholds, perform a logical judgment on the ΔP. i and ΔP solidi A smooth switching is performed between steps, selecting the target power correction value for the submodule within the current control cycle, and updating the state variable z; in step S5), logical judgment is performed based on the state variable z determined in step S4) and the set inner and outer boundary thresholds, specifically including three steps:

[0060] Step S5-1) Calculate the current SOC absolute deviation of each submodule. diffi = |SOC i - SOC ave |;

[0061] Step S5-2) If the state variable z = ΔP determined in step S4) solidi If the system used constant power equalization in the previous control cycle, then the current absolute SOC deviation (SOC) is determined. diffi Is it less than the internal narrow threshold b? If so, then switch the output fine-tuning power correction value ΔP in the current cycle. i And update the state variable z to ΔP i If not, maintain the output forced equalization power correction value ΔP. solidi ;

[0062] Step S5-3) If the state variable z = ΔP determined in step S4) i If the system used a fixed-coefficient balancing method in the previous control cycle, then the current absolute SOC deviation (SOC) is determined. diffi Is it less than the external wide threshold kb? If so, maintain the output fine-tuning power correction value ΔP in the current cycle. i If not, switch the output forced equalization power correction value ΔP. solidi And update the state variable z to ΔP. solidi ;

[0063] S6) Add the target power correction value selected in step S5) to the reference allocated power of the submodule to generate the target reference power P of the submodule. b * In step S6), the target reference power P of the generated submodule is... b * Specifically, it includes two steps:

[0064] Step S6-1) The total rated power P B The power is evenly distributed among the 6N sub-modules of the system to obtain the baseline power distribution P. B / (6N);

[0065] Step S6-2) Add the reference allocated power to the target power correction value selected in step S5) to obtain the target reference power P. b * ;

[0066] S7) Convert the target reference power into a reference current command; the specific method of converting the target reference power into a reference current command in step S7) is as follows: convert the target reference power P b * Divide by the real-time voltage u of the current submodule battery b The reference current command i is obtained. b * ;

[0067] S8) The reference current command is introduced into the inner current loop control of the non-isolated bidirectional DC / DC converter, thereby generating a PWM duty cycle signal D to drive the switching transistor. Step S8) involves two steps:

[0068] Step S8-1) Set the reference current command i b * Compared with the actual sampled battery current i bThe error between them is fed into the dual closed-loop proportional-integral controller of the bidirectional DC / DC converter to calculate the output PWM duty cycle D = (i b *- i b (K) p +K i / s), where K p and K i These are the proportional coefficient and integral coefficient in the proportional-integral controller, respectively, and s is the Laplace operator;

[0069] Step S8-2) Based on the duty cycle, a drive signal is generated using complementary pulse width modulation to control the on and off states of switches T3 and T4, thereby controlling the actual battery current i. b Track the reference current command i b * This enables rapid balancing of single-level SOC.

[0070] This invention takes a modular multilevel converter battery energy storage system as an example. It uses the traditional three-level SOC equalization method based on carrier phase shift dual closed loop, the fixed coefficient method, and the control method of this invention to conduct simulation and experimental verification on the important parameter of SOC equalization time between phases, within bridge arms, and between sub-modules. The principle of its multilevel converter is the same.

[0071] This invention is based on Figure 3 The flowchart shown illustrates a hybrid power control process based on hysteresis logic. A target power correction value is obtained through rigorous logical judgment. This power correction value is then compared with the reference allocation power P. B The target reference power is obtained by adding (6N) together, and finally the target reference power is converted into a reference current command and introduced into the system. Figure 4 The current inner loop control of the battery SOC dual closed-loop equalization control structure of the non-isolated bidirectional DC / DC converter shown is illustrated. This method enables each submodule to independently and directly adjust its actual charging and discharging power based on its deviation from the system average SOC, causing the battery capacity of all submodules in the system to directly align with the system average value, thereby independently completing single-stage SOC fast equalization and completely eliminating the traditional phase-to-phase SOC equalization and upper / lower arm SOC equalization steps.

[0072] Figure 5 This demonstrates the dynamic relationship between the output power correction value ΔP and time t under different control methods during the SOC equalization process, where the integral area represents the equalization energy actually transferred by the system to eliminate the SOC deviation. Figure 5As shown in (a), in the traditional fixed-coefficient method, the output power correction value ΔP continuously decreases as the real-time SOC deviation between batteries decreases during the SOC equalization process. The green area (I) in the figure represents the total energy transferred by this method throughout the equalization period. The fatal flaw of this control mechanism is that in the later stages of equalization, due to the smaller SOC deviation, the output adjustment power becomes extremely weak, resulting in a very lengthy process for eliminating residual small deviations, severely delaying the overall equalization speed of the system. Figure 5 As shown in (b), the constant power method employed in this invention maintains the output power correction value ΔP at a constant, relatively high level determined by the initial large deviation during the SOC equalization process. Within the same timeframe, the total energy transferred by the method of this invention is the sum of the areas of the green region (I) and the blue region (II) in the figure. Comparative analysis shows that within the same equalization cycle, the constant power method outputs an additional transferred energy as shown in region (II) compared to the traditional constant coefficient method. This mechanism fundamentally overcomes the technical deficiency of traditional methods in the later stages of equalization, maintaining maximum energy transfer efficiency throughout the process and fundamentally accelerating the SOC equalization speed.

[0073] from Figure 6 Simulated waveforms and Figure 7 The experimental waveform comparison shows that the SOC equalization time of the multi-level energy storage system using the control method of this invention is much lower than that using the traditional three-level equalization method and the fixed coefficient method, significantly reducing the equalization time between phases, within bridge arms, and between submodules by 67.00% to 79.88%. The benefits include greatly improved system equalization response speed, significantly simplified control architecture, effective avoidance of overcharging or over-discharging problems caused by long-term inconsistent SOC of the battery, and improved battery life and overall operating efficiency of the energy storage system.

[0074] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A constant-power SOC equalization method for a modular multilevel battery energy storage system, characterized in that, Includes the following steps: S1) Real-time acquisition of the State of Charge (SOC) value of each sub-module battery in the energy storage system. i and the total rated power P of the entire battery energy storage system B Where i is the sequence number of the corresponding submodule, i = 1, 2, 3, ..., 6N, and N is the number of submodules connected in series in each bridge arm; S2) According to the SOC i and SOC i_init Calculate the average SOC value of all batteries in the current system. ave and the initial average SOC value at the initial running time of the system. ave_init ; S3) Based on the SOC i SOC ave SOC i_ and SOC ave_init Calculate the fine-tuning power correction value ΔP based on the "fixed coefficient method". i And the forced equalization power correction value ΔP based on the "constant power method". solidi ; S4) Set the inner and outer boundary thresholds of the hysteresis control and determine the equilibrium mode state variable z of the system in the current control cycle; S5) Based on the state variable z determined in step S4) and the set inner and outer boundary thresholds, perform a logical judgment on the ΔP. i and ΔP solidi Smoothly switch between them, select the target power correction value of the submodule in the current control cycle, and update the state variable z; S6) Add the selected target power correction value to the reference allocated power of the submodule to generate the target reference power P of the submodule. b * ; S7) Convert the target reference power into a reference current command; S8) The reference current command is introduced into the current inner loop control of the non-isolated bidirectional DC / DC converter, thereby generating a PWM duty cycle D to drive the switching transistor.

2. The constant-power SOC equalization method for a modular multilevel battery energy storage system according to claim 1, characterized in that, The specific hardware topology connection relationship of each submodule in the energy storage system is as follows: The submodule includes four switching transistors T1, T2, T3 and T4 with anti-parallel diodes, a shared capacitor C, a filter inductor and a battery cell. The main circuit port of the half-bridge structure formed by the series connection of the switching transistors T1 and T2 is connected to the bridge arm of the energy storage system, and the shared capacitor C is connected in parallel across the series branch of the switching transistors T1 and T2. The series connection of the switching transistors T3 and T4 forms the bridge arm of the non-isolated bidirectional DC / DC converter, and the series branch of the switching transistors T3 and T4 is also connected in parallel across the shared capacitor C. The battery cell is connected in series with the filter inductor. One end of the series connection is connected to the midpoint of the series connection of the switching transistors T3 and T4, and the other end is connected to the negative terminal of the shared capacitor C.

3. The constant-power SOC equalization method for a modular multilevel battery energy storage system according to claim 1, characterized in that, In step S3), the fine-tuning power correction value based on the constant coefficient method is: ΔP i = k i × (SOC i - SOC ave The forced equalization power correction value based on the constant power method is: ΔP solidi = k i × (SOC i_init - SOC ave_init ); where k i This is the preset balance ratio coefficient.

4. The constant-power SOC equalization method for a modular multilevel battery energy storage system according to claim 1, characterized in that, In step S4), the specific steps for setting the inner and outer boundary thresholds of the hysteresis control and determining the equilibrium mode state variable z of the system in the current control cycle are as follows: Step S4-1) Set the internal narrow threshold of hysteresis control to b and the external wide threshold to kb; where b is the preset minimum allowable steady-state deviation and k is a shake-proof margin coefficient greater than 1. Step S4-2) During the initial system runtime, calculate the initial absolute SOC deviation of the submodule |SOC i_init - SOC ave_init |; If the initial deviation is greater than or equal to the external wide threshold kb, then assign an initial value z = ΔP to the state variable. solidi This means determining that the system has initially entered a constant power equilibrium mode; if the initial deviation is less than kb, then the initial value of the state variable z = ΔP is assigned. i That is, to determine that the system initially enters a constant coefficient equilibrium mode; Step S4-3) In subsequent control cycles after the system is running normally, directly read the value of the state variable z saved at the end of the previous control cycle; if z = ΔP solidi If z = ΔP, then the system used a constant power equalization method in the previous control cycle; i If so, it is determined that the system used a fixed coefficient equilibrium method in the previous control cycle.

5. The constant-power SOC equalization method for a modular multilevel battery energy storage system according to claim 1, characterized in that, In step S5), the specific steps for performing logical judgments based on the state variable z determined in step S4) and the set inner and outer boundary thresholds are as follows: Step S5-1) Calculate the current SOC absolute deviation of each submodule. diffi = |SOC i - SOC ave |; Step S5-2) If the state variable z = ΔP determined in step S4) solidi If the system used constant power equalization in the previous control cycle, then the current absolute SOC deviation (SOC) is determined. diffi Is it less than the internal narrow threshold b? If so, then switch the output fine-tuning power correction value ΔP in the current cycle. i And update the state variable z to ΔP i If not, maintain the output forced equalization power correction value ΔP. solidi ; Step S5-3) If the state variable z = ΔP determined in step S4) i If the system used a fixed-coefficient balancing method in the previous control cycle, then the current absolute SOC deviation (SOC) is determined. diffi Is it less than the external wide threshold kb? If so, maintain the output fine-tuning power correction value ΔP in the current cycle. i If not, switch the output forced equalization power correction value ΔP. solidi And update the state variable z to ΔP. solidi .

6. The constant-power SOC equalization method for a modular multilevel battery energy storage system according to claim 1, characterized in that, In step S6), the target reference power P of the generated submodule is... b * The specific steps are as follows: Step S6-1) The total rated power P B The power is evenly distributed among the 6N sub-modules of the system to obtain the baseline power distribution P. B / (6N); Step S6-2) Add the reference allocated power to the target power correction value selected in step S5) to obtain the target reference power P. b * .

7. The constant power SOC equalization method for a modular multilevel battery energy storage system according to claim 1, characterized in that, In step S7), the specific method for converting the target reference power into a reference current command is as follows: the target reference power P b * Divide by the real-time voltage u of the current submodule battery b The reference current command i is obtained. b * .

8. A constant-power SOC equalization method for a modular multilevel battery energy storage system according to claim 1, characterized in that, In step S8), the specific steps for introducing the reference current command into the inner current loop control of the non-isolated bidirectional DC / DC converter, thereby generating the drive signal for the switching transistor of the non-isolated bidirectional DC / DC converter, are as follows: Step S8-1) Set the reference current command i b * Compared with the actual sampled battery current i b The error between them is fed into the current inner loop proportional-integral controller of the non-isolated bidirectional DC / DC converter to calculate the output PWM duty cycle D = (i b *- i b (K) p +K i / s), where K p and K i These are the proportional coefficient and integral coefficient in the proportional-integral controller, respectively, and s is the Laplace operator; Step S8-2) Based on the duty cycle D, a drive signal is generated using complementary pulse width modulation to control the on and off states of switches T3 and T4, thereby controlling the actual battery current i. b Track the reference current command i b * This enables rapid balancing of single-level SOC.

Images