Hybrid energy storage system energy management method and system

Abstract

The application discloses a hybrid energy storage system energy management method and system, and relates to the technical field of energy storage system energy management. The method comprises the following steps: collecting output currents of each energy storage converter and estimating physical link impedance values of each converter branch to a DC bus; constructing a physical link decoupling term and superimposing the physical link decoupling term into a droop control model to generate a preliminary voltage reference value; collecting output voltages of each energy storage converter after physical link decoupling, inputting the output voltages into a PI regulator respectively after being subtracted from a set DC bus voltage reference value, calculating a battery-side virtual voltage source compensation term and a super capacitor-side virtual voltage source compensation term, and performing second correction on the preliminary voltage reference value to obtain a final output voltage reference value; and sending the final output voltage reference value into a voltage-current double closed loop controller for modulation to generate a PWM driving signal, so as to control actions of each energy storage converter. The method eliminates high-frequency fluctuation components of the bus voltage under high-frequency and severe fluctuation of the load, and ensures stability of the bus voltage.

Description

Technical Field

[0001] This invention relates to the field of energy management technology for energy storage systems, and specifically to an energy management method and system for hybrid energy storage systems. Background Technology

[0002] Data center power architectures are evolving towards 800V / ±400V high-voltage DC systems to improve end-to-end efficiency. However, due to the surge in power density and frequent switching of operating states of core computing devices, the load exhibits large-amplitude, high-frequency fluctuations, which can easily lead to DC bus voltage instability. A single energy storage device struggles to balance response speed and energy capacity. While energy storage devices such as batteries can provide continuous support, they exhibit lag in response to high-frequency shocks and are prone to accelerated aging; power storage devices such as supercapacitors offer rapid response but have limited capacity. Therefore, introducing a hybrid energy storage system (HESS) composed of batteries and supercapacitors has become a key technological path to mitigate power fluctuations across multiple time scales and maintain DC bus voltage stability.

[0003] Existing hybrid energy storage control strategies mostly adopt a distributed architecture. Introducing an integral droop circuit in the supercapacitor control loop achieves high and low frequency power decomposition, that is, the battery smooths low frequency power fluctuations and the supercapacitor responds to high frequency power fluctuations. However, this scheme has an inherent DC bus steady-state voltage drop, and as the output current of the energy storage unit increases, the deviation of the bus voltage is further aggravated. On this basis, several improvement strategies are proposed to eliminate the problem of DC bus steady-state voltage drop: (1) By collecting the output voltage of the local battery converter, a single voltage compensation circuit is added to the battery control. Although the steady-state drop is solved, the supercapacitor will absorb power from the bus side during the voltage recovery process and is forced to respond to low frequency power fluctuations; (2) Select the output voltage of the supercapacitor converter and use the same single voltage compensation circuit in the battery control. Its implementation relies on the communication network, which increases the communication cost and operation risk of the system; (3) By introducing an integral droop circuit in the battery control loop, the bus steady-state voltage deviation-free control is achieved. However, when this solution is applied to data center scenarios, the bus voltage will oscillate due to the high-frequency and violent fluctuations of the load, making it difficult to meet the stringent power quality requirements of core IT loads.

[0004] However, none of the above methods take into account the physical link between the output of the energy storage converter and the DC bus, resulting in low power distribution accuracy and large transient voltage fluctuations. Summary of the Invention

[0005] To address the shortcomings of existing technologies that fail to consider the physical link between the energy storage converter output and the DC bus, resulting in low power distribution accuracy and large transient voltage fluctuations, this invention proposes an energy management method and system for hybrid energy storage systems. By precisely decoupling the physical link, it provides accurate local converter output voltage observations for autonomous recovery of dual-channel steady-state voltage, thereby achieving high-precision energy management of hybrid energy storage systems and solving the problems existing in the prior art.

[0006] A hybrid energy storage system energy management method is applied to an 800V / ±400V DC power supply system for a data center containing multiple energy storage converters. The method is characterized by the power supply system managing the energy of the hybrid energy storage system based on a distributed droop control model, and includes the following steps:

[0007] Collect the output current of each energy storage converter and estimate the physical link impedance from each energy storage converter branch to the DC bus;

[0008] The output current is multiplied by the physical link impedance to construct a physical link decoupling term; the physical link decoupling term is then superimposed on the droop control model to perform the first correction on the initial output voltage reference value of each energy storage converter, generating a preliminary voltage reference value that can offset physical link interference; wherein, the droop control model is used to realize the high and low frequency power decomposition between each energy storage converter;

[0009] The actual output voltage of each energy storage converter is collected under the control of the initial voltage reference value. The difference between the actual output voltage and the set DC bus voltage reference value is calculated and then the difference is input into their respective independent PI regulators to calculate the virtual voltage source compensation item on the side of each energy storage converter.

[0010] The physical link decoupling term and the virtual voltage source compensation term on each energy storage converter side are synchronously superimposed into the droop control model to make a second correction to the initial voltage reference value, so as to obtain the final output voltage reference value of each energy storage converter after correction.

[0011] The final output voltage reference value is input to the voltage-current dual closed-loop controller for modulation to generate a PWM drive signal to control the operation of each energy storage converter.

[0012] Furthermore, the methods for estimating the physical link impedance value include passive estimation methods based on data tables, passive estimation methods based on hardware measurements, or active estimation methods based on pulse voltage injection.

[0013] Furthermore, the initial output voltage reference value of each energy storage converter includes the initial output voltage reference value of the battery converter and the initial output voltage reference value of the supercapacitor converter; the initial output voltage reference value of the battery converter is determined based on its output current and virtual droop coefficient, and the initial output voltage reference value of the supercapacitor converter is determined based on its output current and virtual droop coefficient, which are expressed as follows:

[0014] ;

[0015] in, v oBiref and v oSCiref They represent the first i The battery inverter and the first i Reference value for the output voltage of a supercapacitor converter; and This represents the corresponding physical link estimate; V ref This indicates the reference value for the DC bus voltage. i oBi and i oSCi They represent the first i The battery inverter and the first i The output current of a supercapacitor converter; k vBi and k vSCi They represent the first i The battery inverter and the first i Virtual droop coefficient of a supercapacitor converter; s It is the Laplace operator.

[0016] Furthermore, the virtual voltage source compensation items on each energy storage converter side include virtual voltage source compensation items on the battery side. and virtual voltage source compensation term on the supercapacitor side , respectively represented as:

[0017] ;

[0018] in, k cpx , k cix These represent the PI element ratio and integral coefficient in the virtual voltage source, respectively. x Indicates B or SC. v oBi and v oSCi Represented as the first i The battery inverter and the first iThe local output voltage of a supercapacitor converter.

[0019] Furthermore, the preliminary voltage reference value is corrected a second time to obtain the final output voltage reference value of each energy storage converter after correction. The calculation process is expressed as follows:

[0020] ;

[0021] in, This indicates the final output voltage reference value of the corrected battery converter. This indicates the final output voltage reference value of the corrected supercapacitor converter.

[0022] This invention also proposes an energy management system for a hybrid energy storage system, comprising:

[0023] The DC bus adopts an 800V / ±400V high-voltage DC architecture.

[0024] Multiple parallel energy storage converter branches are connected to the DC bus via bidirectional DC / DC converters.

[0025] Each energy storage converter branch is independently configured with a control unit; the control unit includes:

[0026] Voltage and current acquisition module, used to acquire local output voltage and output current;

[0027] The physical link estimation module is used to estimate the physical link impedance value from this branch to the DC bus;

[0028] The virtual voltage source compensation module is used to calculate the virtual voltage source compensation terms on each energy storage converter side.

[0029] The voltage reference value generation module is used to superimpose physical link decoupling terms and virtual voltage source compensation terms to generate the final output voltage reference value.

[0030] The PWM modulation module is used to generate drive signals to control the operation of each energy storage converter based on the final output voltage reference value.

[0031] Furthermore, the multiple energy storage converters are divided into a first group and a second group; the positive output of the first group of energy storage converters is connected to the +400V bus, and the negative output is connected to the system neutral line; the negative output of the second group of energy storage converters is connected to the -400V bus, and the positive output is connected to the system neutral line; the +400V bus and the -400V bus form an 800V DC bus in series.

[0032] Furthermore, the physical link estimation module pre-stores the physical link impedance value using an offline calibration method, or calculates the physical link impedance value in real time using an online estimation method.

[0033] Furthermore, the PI regulator parameters in the virtual voltage source compensation module are set independently according to the dynamic response characteristics of the energy storage medium.

[0034] Furthermore, the system is connected to the AC power grid via a solid-state transformer and is connected to wind power and photovoltaic new energy power generation units to jointly form a hybrid power supply architecture for the data center.

[0035] This invention provides an energy management method for a hybrid energy storage system, which has the following beneficial effects:

[0036] This invention introduces a physical link decoupling term to counteract the coupling effect of transmission channel parameters at the control level, achieving precise frequency division between different energy storage media, accurate power allocation among multiple converters within the same energy storage medium, and reduced bus transient voltage fluctuation amplitude. Secondly, a virtual voltage source based on a PI element is added to the control loop of the converters for both energy storage media. This dual-channel mechanism ensures no steady-state voltage drop on the bus while guaranteeing that the supercapacitor does not participate in low-frequency power regulation during the bus voltage recovery process, avoiding cross-medium power misinterpretation and eliminating the need for a communication network. Finally, relying on the high-frequency disturbance suppression characteristics of the supercapacitor-side integral element, the high-frequency fluctuation component of the bus voltage under severe high-frequency load fluctuations is eliminated, ensuring bus voltage stability. Attached Figure Description

[0037] Figure 1 This is a schematic diagram of an 800V / ±400V DC power supply system for a data center containing a multi-energy storage converter (HESS) according to an embodiment of the present invention.

[0038] Figure 2 This is a simplified schematic diagram of an 800V / ±400V DC power supply system with multiple energy storage converters (HESS) in an embodiment of the present invention.

[0039] Figure 3 This is a control block diagram of the battery module in an embodiment of the present invention;

[0040] Figure 4 This is a control block diagram of the supercapacitor module in an embodiment of the present invention.

[0041] Figure 5 This is a schematic diagram of the energy management method for a hybrid energy storage system in an embodiment of the present invention. Detailed Implementation

[0042] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0043] This invention proposes an energy management method for hybrid energy storage systems. This method introduces a physical link decoupling term to counteract the coupling effects of transmission channel parameters at the control layer, achieving precise frequency division between different energy storage media, accurate power allocation among multiple converters within the same energy storage medium, and reduced bus transient voltage fluctuations. Secondly, a virtual voltage source based on a PI circuit is added to the control loops of the converters for both energy storage media. This dual-channel mechanism ensures no steady-state voltage drop on the bus while guaranteeing that the supercapacitor does not participate in low-frequency power regulation during bus voltage recovery, avoiding cross-medium power misinterpretation and eliminating reliance on communication networks. Finally, relying on the high-frequency disturbance suppression characteristics of the supercapacitor-side integral circuit, the high-frequency fluctuation component of the bus voltage under severe high-frequency load fluctuations is eliminated, ensuring bus voltage stability.

[0044] The structure of a data center 800V / ±400V DC power supply system with HESS multi-energy storage converter is as follows: Figure 1 As shown, the 10kV AC grid undergoes AC / DC conversion via a solid-state transformer, outputting a ±400V DC bus voltage. Wind power, photovoltaic, and other new energy sources are connected to the DC bus voltage via unidirectional AC / DC and DC / DC converters, respectively. The HESS converter, composed of batteries and supercapacitors and containing multiple energy storage devices, is connected to the ±400V bus via a bidirectional DC / DC converter, undertaking functions such as smoothing power fluctuations, maintaining bus voltage stability, and compensating for power shortages. In the power supply process, the DC bus power is branched by the distribution unit and sent to the power supply unit. After being stepped down by two levels of isolation (including the high-voltage intermediate bus converter HVIBC and the low-voltage intermediate bus converter LVIBC) to achieve voltage adaptation, the voltage regulator VR precisely outputs the adapted operating voltage, ensuring the reliable and stable operation of core IT loads such as the central processing unit and memory. Figure 2 A simplified schematic diagram of an 800V / ±400V DC power supply system, configured with m A parallel battery converter branch, and n A parallel supercapacitor converter branch. This diagram only shows the connection topology between the +400V bus and the energy storage converter. The -400V bus structure is similar; simply connect the negative output of the energy storage converter to the -400V bus and the positive output to the system neutral line. R lBm Indicates the first m The physical link value from the battery converter output to the DC bus; i oBm Indicates the first m The output current of the battery converter; v oBm Indicates the first m The output voltage of the battery converter; C Bm Indicates the first mThe capacitor of the battery converter; L Bm Indicates the first m The inductor of the battery converter; S B1m Indicates the first m Switch No. 1 of the battery converter; S B2m Indicates the first m Switch No. 2 of the battery converter; i LBm Indicates the first m The inductor current of the battery converter; v Bm Indicates the first m Battery voltage, if the above symbols are... m A value of 1 indicates the first battery converter. When the subscript is SC, it corresponds to the parameter symbol on the supercapacitor side.

[0045] This method achieves high-precision energy management of hybrid energy storage systems by precisely decoupling the physical links and providing accurate local converter output voltage observations for autonomous recovery of dual-channel steady-state voltage. The overall architecture is distributed, with each converter only collecting local electrical quantities, eliminating the need for interconnection via communication networks. Figure 5 As shown, the specific steps include:

[0046] S1. Collect the output current of each energy storage converter and the reference value of the given DC bus voltage through sensors; use physical link estimation methods (including passive methods based on data tables or standard models, passive methods based on hardware measurements, and active estimation techniques such as pulse voltage injection) to calculate the estimated physical link value from each energy storage converter branch to the DC bus online or offline. and .

[0047] S2. Physical Link Decoupling Modeling: Multiply the obtained converter output current with the corresponding physical link estimate to construct a physical link decoupling term, which is then superimposed on the traditional droop control model to complete the physical link decoupling.

[0048] After introducing physical link decoupling, the reference value for the output voltage of the energy storage converter is:

[0049] (1)

[0050] in, v oBiref and v oSCiref They represent the first i The battery inverter and the first i Reference value for the output voltage of a supercapacitor converter; and This represents the corresponding physical link estimate; s It is the Laplace operator; k vBi and k vSCi Indicates the first i The battery inverter and the first i Virtual droop coefficient of a supercapacitor converter; i oBi and i oSCi Indicates the first i The battery inverter and the first i The output current of a supercapacitor converter; V ref This indicates the reference value for the DC bus voltage.

[0051] After introducing the physical link decoupling term, the ideal frequency domain current distribution relationship of a HESS with multiple energy storage converters is as follows:

[0052] (2)

[0053] in, i oB This represents the sum of the output currents of all battery converters; i oSC This represents the sum of the output currents of all supercapacitor converters; i o This represents the total output current of a HESS containing multiple energy storage converters; Δ R lBi With Δ R lSCi These represent the physical circuit decoupling errors:

[0054] (3)

[0055] When Δ R lBi With Δ R lSCi When the value is 0, that is, when the physical link is precisely decoupled, equation (2) is simplified to the ideal frequency domain current distribution relationship, restoring the precise frequency division between different energy storage media.

[0056] For a HESS converter with multiple energy storage units, after introducing a physical link decoupling term, the converter output current relationship becomes:

[0057] (4)

[0058] Precise decoupling of physical links ensures accurate power distribution among multiple converters within the same energy storage medium.

[0059] After introducing the physical link decoupling term, the bus voltage expression can be written as:

[0060] (5)

[0061] When the line impedance is accurately estimated, the output voltage of the energy storage converter is equal to the DC bus voltage. The bus voltage fluctuation is no longer affected by the physical link, but mainly depends on the droop element in the control strategy. Combining equation (1), it can be seen that due to the integral element in the control law, the voltage drop caused by the current during high-frequency disturbances will approach zero. Therefore, the degree of fluctuation of the DC bus voltage depends only on the performance of the PI controller itself.

[0062] In steady state, the DC bus voltage is:

[0063] (6)

[0064] The above equation shows that the steady-state voltage of the bus has recovered to the ideal condition, but there is still an inherent voltage drop caused by droop control.

[0065] S3, Dual-channel steady-state voltage autonomous recovery modeling.

[0066] S3.1. Collect the actual output voltages of the battery and supercapacitor converter after physical link decoupling, and compare them with the rated reference voltage of the DC bus. V ref The difference is calculated and input into the respective independent PI regulators to obtain the dual-channel virtual voltage source compensation term Δ. V Bi With Δ V SCi .

[0067] (7)

[0068] In the above formula k cpx , k cix These represent the PI element ratio and integral coefficient in the virtual voltage source, respectively. x (Indicates B or SC); , They are the first i The battery inverter and the first i The output voltage of a supercapacitor converter.

[0069] S3.2. Simultaneously superimpose the physical link decoupling term and the virtual voltage source compensation term into the droop control model to calculate the corrected final output voltage reference value of the energy storage converter. and .

[0070] To eliminate steady-state voltage dips on the DC bus and avoid cross-dielectric power misinterpretation caused by single-ended compensation, this invention synchronously adds a virtual voltage source compensation term Δ based on a PI element to the control loops of both the battery and supercapacitor converters. V Bi With Δ V SCi A dual-channel steady-state voltage self-recovery mechanism was constructed. This mechanism does not rely on a communication network, and while achieving no steady-state voltage drop at the bus, it ensures that the supercapacitor does not participate in low-frequency power regulation during the voltage recovery phase, fundamentally cutting off the physical path of cross-dielectric power crosstalk. The control block diagrams are as follows: Figure 3 and Figure 4 As shown, the final converter output voltage reference value is:

[0071] (8)

[0072] in, This indicates the final output voltage reference value of the corrected battery converter. This represents the final output voltage reference value of the corrected supercapacitor converter; Δ V Bi With Δ V SCi The value is obtained by subtracting the local output voltage of the converter from the reference voltage, and the difference is obtained through a PI circuit.

[0073] S4. The corrected voltage reference value is sent to the voltage-current dual closed-loop controller for modulation to generate a PWM drive signal, which controls the operation of each energy storage converter to achieve precise power distribution and stable bus voltage control.

[0074] This invention addresses the problems of physical link interference and power crosstalk caused by single-sided compensation in traditional hybrid energy storage control strategies. It proposes a high-precision energy management strategy based on physical link decoupling and dual-channel steady-state voltage autonomous recovery. By introducing a physical link decoupling term, precise frequency division between different energy storage media, accurate power allocation among multiple converters within the same energy storage medium, and reduced bus transient voltage fluctuations are achieved, ensuring accurate observation of the local converter output voltage during the steady-state voltage autonomous recovery phase. Simultaneously, this invention innovatively constructs a dual-channel PI virtual voltage source simultaneously on both sides of the battery and the supercapacitor. This ensures zero bus steady-state voltage drop while maintaining a zero relative voltage difference across the supercapacitor during the low-frequency transient process of bus voltage recovery. This completely cuts off the physical path for the supercapacitor to absorb low-frequency power, eliminating cross-medium power crosstalk during voltage recovery, and requiring no reliance on a communication network throughout the entire process.

[0075] Based on the same inventive concept, this invention also proposes an energy management system for a hybrid energy storage system, comprising:

[0076] The DC bus adopts an 800V / ±400V high-voltage DC architecture; multiple energy storage converters are divided into a first group and a second group; the positive output of the first group of energy storage converters is connected to the +400V bus, and the negative output is connected to the system neutral line; the negative output of the second group of energy storage converters is connected to the -400V bus, and the positive output is connected to the system neutral line; the +400V bus and the -400V bus form an 800V DC bus in series.

[0077] Multiple parallel battery converter branches are connected to the DC bus via bidirectional DC / DC converters.

[0078] Multiple parallel supercapacitor converter branches are connected to the DC bus via bidirectional DC / DC converters.

[0079] Each converter branch has an independently configured control unit, which includes:

[0080] The voltage and current acquisition module is used to acquire local output voltage and output current.

[0081] The physical link estimation module is used to estimate the physical link impedance value from this branch to the DC bus; wherein, the physical link impedance value is pre-stored by offline calibration or the physical link impedance value is calculated in real time by online estimation.

[0082] The virtual voltage source compensation module is used to calculate the PI compensation term on the battery side or the supercapacitor side; the PI regulator parameters in the virtual voltage source compensation module are set independently according to the dynamic response characteristics of the energy storage medium.

[0083] The voltage reference value generation module is used to superimpose physical link decoupling terms and virtual voltage source compensation terms to generate the final output voltage reference value.

[0084] The PWM modulation module is used to generate a drive signal based on the final output voltage reference value.

[0085] The system is connected to the AC power grid via a solid-state transformer and is also connected to wind power and photovoltaic new energy power generation units, together forming a hybrid power supply architecture for the data center.

[0086] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A hybrid energy storage system energy management method applied to a 800V / ±400V DC power supply system of a data center with multiple energy storage converters, characterized in that, The power supply system manages the energy of the hybrid energy storage system based on a distributed droop control model, including the following steps: Collect the output current of each energy storage converter and estimate the physical link impedance from each energy storage converter branch to the DC bus; The output current is multiplied by the physical link impedance to construct a physical link decoupling term; the physical link decoupling term is then superimposed on the droop control model to perform the first correction on the initial output voltage reference value of each energy storage converter, generating a preliminary voltage reference value that can offset physical link interference; wherein, the droop control model is used to realize the high and low frequency power decomposition between each energy storage converter; The actual output voltage of each energy storage converter is collected under the control of the initial voltage reference value. The difference between the actual output voltage and the set DC bus voltage reference value is calculated and then the difference is input into their respective independent PI regulators to calculate the virtual voltage source compensation item on the side of each energy storage converter. The physical link decoupling term and the virtual voltage source compensation term on each energy storage converter side are synchronously superimposed into the droop control model to make a second correction to the initial voltage reference value, so as to obtain the final output voltage reference value of each energy storage converter after correction. The final output voltage reference value is input to the voltage-current dual closed-loop controller for modulation to generate a PWM drive signal to control the operation of each energy storage converter.

2. The energy management method of a hybrid energy storage system according to claim 1, wherein, The methods for estimating the physical link impedance include passive estimation methods based on data tables, passive estimation methods based on hardware measurements, or active estimation methods based on pulse voltage injection.

3. The energy management method for a hybrid energy storage system according to claim 1, characterized in that, The initial output voltage reference values ​​for each energy storage converter include the initial output voltage reference values ​​for the battery converter and the supercapacitor converter. The initial output voltage reference value for the battery converter is determined based on its output current and virtual droop coefficient, and the initial output voltage reference value for the supercapacitor converter is determined based on its output current and virtual droop coefficient, respectively expressed as follows: ； in, v oBiref and v oSCiref They represent the first i The battery inverter and the first i Reference value for the output voltage of a supercapacitor converter; and This represents the corresponding physical link estimate; V ref This indicates the reference value for the DC bus voltage; i oBi and i oSCi They represent the first i The battery inverter and the first i The output current of a supercapacitor converter; k vBi and k vSCi They represent the first i The battery inverter and the first i Virtual droop coefficient of a supercapacitor converter; s It is the Laplace operator.

4. The energy management method for a hybrid energy storage system according to claim 3, characterized in that, The virtual voltage source compensation items on each energy storage converter side include the virtual voltage source compensation items on the battery side. and virtual voltage source compensation term on the supercapacitor side , respectively represented as: ； in, k cpx , k cix These represent the PI element ratio and integral coefficient in the virtual voltage source, respectively. x Indicates B or SC. v oBi and v oSCi Represented as the first i The battery inverter and the first i The local output voltage of a supercapacitor converter.

5. The energy management method for a hybrid energy storage system according to claim 4, characterized in that, The second correction of the initial voltage reference value yields the final output voltage reference value for each energy storage converter after correction. The calculation process is as follows: ； in, This indicates the final output voltage reference value of the corrected battery converter. This indicates the final output voltage reference value of the corrected supercapacitor converter.

6. An energy management system for implementing a hybrid energy storage system as described in claim 1, characterized in that, include: The DC bus adopts an 800V / ±400V high-voltage DC architecture. Multiple parallel energy storage converter branches are connected to the DC bus via bidirectional DC / DC converters. Each energy storage converter branch is independently configured with a control unit; the control unit includes: Voltage and current acquisition module, used to acquire local output voltage and output current; The physical link estimation module is used to estimate the physical link impedance value from this branch to the DC bus; The virtual voltage source compensation module is used to calculate the virtual voltage source compensation terms on each energy storage converter side. The voltage reference value generation module is used to superimpose physical link decoupling terms and virtual voltage source compensation terms to generate the final output voltage reference value. The PWM modulation module is used to generate drive signals to control the operation of each energy storage converter based on the final output voltage reference value.

7. The energy management system for a hybrid energy storage system according to claim 6, characterized in that, Multiple energy storage converters are divided into a first group and a second group; the positive output of the first group of energy storage converters is connected to the +400V bus, and the negative output is connected to the system neutral line; the negative output of the second group of energy storage converters is connected to the -400V bus, and the positive output is connected to the system neutral line; the +400V bus and the -400V bus form an 800V DC bus in series.

8. The energy management system for a hybrid energy storage system according to claim 6, characterized in that, The physical link estimation module can either pre-store the physical link impedance value using an offline calibration method or calculate the physical link impedance value in real time using an online estimation method.

9. The energy management system for a hybrid energy storage system according to claim 6, characterized in that, The PI regulator parameters in the virtual voltage source compensation module are set independently according to the dynamic response characteristics of the energy storage medium.

10. The energy management system for a hybrid energy storage system according to claim 6, characterized in that, The system is connected to the AC power grid via a solid-state transformer and is also connected to wind power and photovoltaic new energy power generation units, together forming a hybrid power supply architecture for the data center.

Images