An all-electric propulsion ship parallel energy storage battery state of charge equalization hierarchical coordination control method

Abstract

This invention discloses a hierarchical coordinated control method for the state-of-charge (SOC) balancing of parallel energy storage batteries in an all-electric propulsion ship. The method includes the following steps: acquiring parameters of multiple energy storage battery units and their corresponding connected bidirectional DC-DC converters; calculating the SOC and depth of discharge of the battery units using the Coulomb integral method; iteratively calculating the estimated global average SOC and average depth of discharge of the energy storage power system using a dynamic consensus algorithm; dynamically calculating the adaptive droop coefficient of each energy storage battery unit using a nonlinear formula containing an error function; obtaining the global average value of the information mode factor using a consensus algorithm; generating a voltage compensation amount based on the bus voltage deviation and the information mode factor deviation; correcting the primary voltage control command; and generating the final voltage control command, which is used by the bidirectional DC-DC converter corresponding to each energy storage battery unit to achieve precise power distribution of the energy storage battery units and recovery of the bus voltage.

Description

Technical Field

[0001] This invention belongs to the field of energy storage technology and relates to a hierarchical coordinated control method for the balanced state of charge of parallel energy storage batteries in all-electric propulsion ships. Background Technology

[0002] With the rapid development of renewable energy, the stability and efficiency of new energy ship battery systems are crucial as a key technology for energy conversion and storage. In these systems, batteries have become the primary energy storage medium due to their high energy density and long lifespan. However, during long-term operation, individual differences and uneven charging and discharging can easily lead to inconsistencies in the System-on-Chip (SoC) within the battery pack, affecting battery performance and lifespan. Traditional equalization control methods suffer from slow response, low efficiency, and the risk of fire due to overcharging and over-discharging, making it difficult to meet the demands of modern energy storage systems for fast and accurate SoC equalization. Therefore, designing an equalization control method that is fast-responding, efficient, avoids overcharging or over-discharging, and has minimal impact on bus voltage stability is essential. Summary of the Invention

[0003] To address the aforementioned problems, the technical solution adopted by this invention is: a hierarchical coordinated control method for the state-of-charge balance of parallel energy storage batteries in an all-electric propulsion ship, comprising the following steps:

[0004] S1: Obtain the parameters of multiple energy storage battery units and their corresponding connected bidirectional DC-DC converters, and establish a distributed communication network based on the consensus of energy storage battery units; S2: Obtain the initial SoC, capacitance, charge / discharge efficiency, and current of the energy storage battery cell in the distributed communication network, and use the coulomb integral method to calculate the state of charge and depth of discharge of the energy storage battery cell. S3: Based on energy storage battery cells, the state of charge and depth of discharge of adjacent energy storage battery cells are obtained through low-bandwidth communication links. The global average state of charge and average depth of discharge of the energy storage power system are obtained by iterative calculation using a dynamic consensus algorithm. S4: Based on the deviation between the state of charge of each energy storage battery cell and the global average state of charge, and the deviation between the depth of discharge of each energy storage battery cell and the global average depth of discharge, the adaptive droop coefficient of each energy storage battery cell is dynamically calculated using a nonlinear formula that includes an error function. S5: Based on the adaptive droop coefficient of each energy storage battery cell, a primary voltage control command is generated and executed by the converter corresponding to each energy storage battery cell to achieve the initial power allocation of each energy storage battery cell and the initial SoC balance. S6: After the initial power allocation and initial balancing of each energy storage battery cell, based on the secondary control strategy, the local power allocation influence coefficient and information mode factor of each energy storage battery cell are first calculated, and the average value of the information mode factor is obtained through the consensus algorithm. Then, the voltage compensation amount is generated by integrating the quotient of the average value of the information mode factor and the power allocation influence coefficient minus the bus voltage reference value. Finally, the primary voltage control command is corrected to generate the final voltage control command, thereby realizing the accurate power allocation of the energy storage battery cell and the recovery of the bus voltage.

[0005] Furthermore, the expression for calculating the state of charge of the energy storage battery cell using the Coulomb integral method is as follows; (1) In equation (1), SoC i (0) indicates BSU i The initial SoC, η i BSU i The charge and discharge efficiency, i bati ( τ ) indicates BSU i The charging and discharging current, C bati BSU i The capacity.

[0006] Furthermore, the expression for the dynamic consensus algorithm is: (2) In formula (2) X i =[SoC i , φ i ] indicates BSU i The calculated data, N i It is the set of its neighboring nodes. a ij It is a communication weighting coefficient; X avg _ i ( k )and X avg _ i ( k +1) represent iterations. k and k Local average estimate at +1; BSU i and BSU j The cumulative error value between them is used D ij ( k) indicates that it is initialized to D ij (0) = 0, while binary variables δ ij ∈{0,1} indicates whether there is a communication link between the corresponding nodes.

[0007] Furthermore, when the current of the energy storage battery cell is less than 0, the formula for calculating the droop factor is: (3) In equation (3), C bati , max For BSU i Maximum capacity, C bati For BSU i Rated capacity, R v0 This is the initial droop coefficient. erf [ k ρ SoC i ]、 erf [ k ρ DoD i ] is the error function. k For convergence speed, ρ For acceleration factor, SoC i It's BSU i SoC, DoD i It's BSU i Discharge depth, ΔDoD i It is the difference between the depth of discharge and the average DoD, SoC a and DoD a They are BSU i The average state of charge and average depth of discharge.

[0008] Furthermore, when the current of the energy storage battery cell is ≥0, the formula for calculating the droop coefficient is: (4) In equations (3) and (4), (5) (6) In equation (6), ε For convergence accuracy; (7) (8) (9) Furthermore, primary voltage control commands are generated. V dci ,for: (10) Utilizing generated voltage compensation amount δv i The primary voltage control command is modified to simultaneously achieve precise power distribution and bus voltage recovery and stabilization. The modified primary voltage control command is expressed as follows: (11) Furthermore, the secondary control strategy employs a single integral controller to simultaneously ensure accurate current distribution and bus voltage recovery. The specific method is as follows: (12) In equation (12), V dc The bus voltage reference value is set to 750V. avg for i The average value; λ i For the first i The power distribution influence coefficient of the converter, where avg The expression is shown below. (13) In equation (13), i It is the first i The information mode factor of each converter, and i = λ i × V dci ,in λ i The expression is shown below. (14) In equation (14) P pui The expression is as follows: (15) In equations (14) and (15), P pui For unit output power, R viFor adaptive droop coefficient, b It is a constant. R vi_max It is the maximum value of the adaptive droop coefficient under the set relationship curve between balancing accuracy and balancing speed. P dci For BSU i 'output power' P ratei For BSU i Rated output power, b It is a constant, and its function is to prevent when R vi P dci = R vi_max P ratei When the denominator is 0, R vi P dci / R vi_max P ratei express R vi P dci and R vi_max P ratei The proportion is to i Limited to the range of 0-1, and must satisfy 1- R vi P dci / b R vi_max.。

[0009] A new energy ship controls its power system using a hierarchical coordinated control method based on the state-of-charge balance of parallel energy storage batteries in any of the all-electric propulsion ships.

[0010] A hierarchical coordinated control device for the state-of-charge balancing of parallel energy storage batteries in an all-electric propulsion vessel includes: The first acquisition module acquires parameters of multiple energy storage battery units and their corresponding connected bidirectional DC-DC converters, and establishes a distributed communication network based on the consensus of the energy storage battery units. The second acquisition module acquires the initial state of charge, capacitance, charge / discharge efficiency, and current of the energy storage battery cells in the distributed communication network, and uses the coulomb integral method to calculate the state of charge and depth of discharge of the energy storage battery cells. The first calculation unit: Based on the energy storage battery unit (BSU), it exchanges state of charge and depth of discharge information with adjacent units through a low-bandwidth communication link. Then, based on the state of charge and depth of discharge of all energy storage battery units, it uses a dynamic consensus algorithm to iteratively calculate and obtain the global average state of charge and average DoD estimate of the energy storage power system. The second calculation unit dynamically calculates the adaptive droop coefficient of each energy storage battery cell based on the deviation between the state of charge of each energy storage battery cell and the global average state of charge, and the deviation between the depth of discharge of each energy storage battery cell and the global average depth of discharge, using a nonlinear formula that includes an error function. Main control layer: Used to generate primary voltage control commands based on the adaptive droop coefficient of each energy storage battery cell, and execute them through the corresponding converter of each energy storage battery cell to achieve the initial power allocation of each energy storage battery cell and the initial SoC balancing; Secondary control layer: After the initial power allocation and initial balancing of each energy storage battery cell, based on the secondary control strategy, the local power allocation influence coefficient and information mode factor of each energy storage battery cell are first calculated, and the average value of the information mode factor is obtained through the consensus algorithm. Then, the voltage compensation amount is generated by integrating the quotient of the average value of the information mode factor and the power allocation influence coefficient minus the bus voltage reference value. Finally, the primary voltage control command is corrected to generate the final voltage control command, so as to realize the accurate power allocation of the energy storage battery cell and the recovery of the bus voltage.

[0011] This invention proposes a hierarchical coordinated control method for balancing the state of charge (SoC) of parallel energy storage batteries in all-electric propulsion ships. By designing an adaptive droop coefficient function to dynamically adjust the battery's charge and discharge rates, it achieves SoC balancing within the battery pack. The core of this technology lies in combining SoC balancing control with a dynamic consistency algorithm to propose a battery SoC balancing control strategy suitable for electric propulsion ships. By constructing a hierarchical control structure, it achieves multi-objective coordinated control, including SoC balancing, precise current distribution, and DC bus voltage stability. The main contributions of this chapter are as follows: 1) In the primary control layer, fast SoC equalization is achieved by introducing a cumulative error function with dual adjustment factors and dynamically adjusting the droop coefficient.

[0012] 2) In the secondary control layer, an information mode factor and a single integral controller are combined to ensure accurate current distribution and stable regulation of bus voltage while compensating for the influence of line impedance.

[0013] 3) In the communication layer, information exchange between adjacent communications is established based on the dynamic consensus algorithm, thereby improving the fault tolerance of the communication network and the stability of system operation. Attached Figure Description

[0014] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0015] Figure 1 This is a block diagram of the new energy ship battery power system of the present invention; Figure 2 This is a topology diagram of the control method proposed in this invention; Figure 3 This is a communication expansion diagram between distributed batteries in this invention; Figure 4 In the charging mode of this invention R vi Follow k and ε A surface plot showing the changes in value; Figure 5 In the discharge mode of this invention R vi Follow k and ε A surface plot showing the changes in value; Figure 6 The diagram shows the charge-discharge droop curves of the BSUs of this invention; where (a) is during discharge; and (b) is during charging. Figure 7 The simulation results for propulsion and auxiliary load fluctuations in this invention are shown, where (a) output power, (b) SoC of the lithium battery, (c) output current, and (d) DC bus voltage. Figure 8 The simulation results for SoC equalization speed comparison in this invention are shown, where (a) is the proposed strategy; and (b) are references. [1] Strategy; (c) References [2] Strategy; (d) References [3] Strategy; Figure 9 The simulation results under the plug-and-play test in this invention are shown, where (a) output power; (b) lithium battery SoC; (c) output current; and (d) DC bus voltage. Figure 10 The simulation results under photovoltaic fluctuations and communication anomalies in this invention are shown, where (a) output power, (b) lithium battery SoC, (c) output current, and (d) DC bus voltage. Figure 11The simulation results for the ferry navigation scenario in this invention are shown, where (a) output power, (b) lithium battery SoC, (c) output current, and (d) DC bus voltage. Figure 12 The simulation results are shown in the battery scalability scenario of the all-electric propulsion ship in this invention, where (a) output power, (b) SoC of lithium battery, (c) output current, and (d) DC bus voltage. Figure 13 The experimental results under propulsion and auxiliary load fluctuations in this invention are shown, wherein (a) the SoC of the lithium battery; (b) the output current; and (c) the DC bus voltage. Figure 14 The following are experimental results for different energy storage capacities in this invention, where (a) the SoC of the lithium battery; (b) the output current; and (c) the DC bus voltage. Detailed Implementation

[0016] It should be noted that, unless otherwise specified, the embodiments and features in the embodiments of the present invention can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and embodiments.

[0017] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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 a part of the embodiments of the present invention, and not all of the embodiments. The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit the present invention or its application or use. 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.

[0018] A hierarchical coordinated control method for the state-of-charge balance of parallel energy storage batteries in an all-electric propulsion ship includes the following steps: S1: Obtain the parameters of multiple energy storage battery units and their corresponding connected bidirectional DC-DC converters, and establish a distributed communication network based on the consensus of energy storage battery units; S2: Obtain the initial SoC, capacitance, charge / discharge efficiency, and current of the energy storage battery cell in the distributed communication network, and use the coulomb integral method to calculate the state of charge and depth of discharge of the energy storage battery cell. S3: Based on battery storage units (BSUs), the state of charge and depth of discharge of adjacent battery storage units are obtained through a low-bandwidth communication link. The global average SoC and average DoD of the energy storage power system are obtained by iterative calculation using a dynamic consensus algorithm. S4: Based on the deviation between the SoC of each energy storage cell and the global average SoC, and the deviation between the DoD of each energy storage cell and the global average DoD, through an error function ( erf The nonlinear formula is used to dynamically calculate the adaptive droop coefficient for each energy storage battery cell. R vi ); S5: Generate primary voltage control commands based on the adaptive droop coefficient of each energy storage battery cell. V dci And through the converter corresponding to each energy storage battery unit, the initial power allocation of each energy storage battery unit and the initial balancing of the SoC are achieved; S6: After the initial power allocation of each energy storage battery cell and the initial balancing of the SoC, calculate the local power allocation influence coefficient of each energy storage battery cell. λ i With information modality factor i The average value of the information mode factor is obtained through a consensus algorithm. avg Then, based on the bus voltage deviation and the information mode factor deviation, a voltage compensation amount is generated to adjust the primary voltage control command ( V dc - i dci R vi The system corrects the voltage and generates a final voltage control command, which is then applied to the bidirectional DC-DC converter corresponding to each energy storage battery unit to achieve precise power distribution and bus voltage recovery for the energy storage battery units.

[0019] The generated final voltage control command acts on the bidirectional DC-DC converter, driving the power switching devices; when SoC i When ≈0 is not true, the system executes steps two through six in a loop to achieve continuous and dynamic SoC balancing and system stability control.

[0020] Steps S1 / S2 / S3 / S4 / S5 / S6 are executed sequentially. Furthermore, the expression for calculating the state of charge of the energy storage battery cell using the Coulomb integral method is as follows; (1) In equation (1), SoC i (0) indicates BSU i The initial SoC, η i BSU iThe charge and discharge efficiency, i bati ( τ ) indicates BSU i The charging and discharging current, i express Furthermore, the expression for the dynamic consensus algorithm is: (2) in X i =[SoC i , φ i ] indicates BSU i The calculated data, N i It is the set of its neighboring nodes. a ij It is a communication weighting coefficient; X avg _ i ( k )and X avg _ i ( k +1) represent iterations. k and k Local average estimate at +1; BSU i and BSU j The cumulative error value between them is used D ij ( k ) indicates that it is initialized to D ij (0) = 0, while binary variables δ ij ∈{0,1} indicates whether a communication link exists between the corresponding nodes. The algorithm needs to choose the appropriate values. a ij To meet stability requirements.

[0021] Furthermore, when the current of the energy storage battery cell is less than 0, the formula for calculating the droop factor is: (3) In equation (3), C bati , max For BSU i Maximum capacity, C bati For BSU i Rated capacity, R v0 This is the initial droop coefficient. erf [ k ρ SoC i ]、 erf [ k ρ DoD i ] is the error function. k For convergence speed, ρ For acceleration factor, SoC i It's BSU i SoC, DoD i It's BSU i Discharge depth, ΔDoD i It is the difference between the depth of discharge and the average DoD, SoC a and DoD a They are BSU i Average SoC and average depth of discharge.

[0022] Furthermore, when the current of the energy storage battery cell is greater than 0, the formula for calculating the droop factor is: (4) In equations (3) and (4), (5) (6) In equation (6), ε For convergence accuracy; (7) (8) (9) (10) Furthermore, primary voltage control commands are generated. V dci ,for: (11) When the bus voltage deviation and the mode information factor deviation generate voltage compensation amount δv i The primary voltage control command is modified to simultaneously achieve precise power distribution and bus voltage recovery and stabilization. The modified primary voltage control command is expressed as follows: (16) Furthermore, the secondary control strategy employs a single integral controller to simultaneously ensure accurate current distribution and bus voltage recovery. The specific method is as follows: (12) In equation (12), V dc The bus voltage reference value is set to 750V. avg for i The average value; λ i For the first i The power distribution influence coefficient of the converter, where avg The expression is shown below. (13) In equation (13), i It is the first i The information mode factor of each converter, and i = λ i × V dci ,in λ i The expression is shown below. (14) In equation (14) P pui The expression is as follows: (15) In equation (15), P pui For unit output power, R vi For adaptive droop coefficient, b It is a constant. R vi_max This refers to the maximum value of the adaptive droop coefficient under the set equalization accuracy and equalization speed curve. P dci For BSU i 'output power' P ratei For BSU i Rated output power, b It is a constant. b =2 is to prevent R vi P dci = R vi_max P ratei When the denominator is 0. Assume... Pratei =15kW and R vi_max =4Ω. R vi P dci / R vi_max P ratei express R vi P dci and R vi_max P ratei The proportion. We need to... λ i Limited to between 0 and 1, therefore 1- is required. P pui R vi / bR vi_max This way, the output power of the unit converter will not be too high.

[0023] A hierarchical coordinated control method for the state-of-charge balance of parallel energy storage batteries in any type of all-electric propulsion ship is proposed to control the power system.

[0024] A hierarchical coordinated control device for the state-of-charge balancing of parallel energy storage batteries in an all-electric propulsion vessel includes: The first acquisition module acquires parameters of multiple energy storage battery units and their corresponding connected bidirectional DC-DC converters, and establishes a distributed communication network based on the consensus of the energy storage battery units. The second acquisition module acquires the initial SoC, capacitance, charge / discharge efficiency, and current of the energy storage battery cells in the distributed communication network, and uses the coulomb integral method to calculate the state of charge and depth of discharge of the energy storage battery cells. First calculation unit: Based on the energy storage battery unit (BSU), it exchanges SoC and DoD information with adjacent units through a low-bandwidth communication link. Then, based on the state of charge and depth of discharge of all energy storage battery units, it uses a dynamic consensus algorithm to iteratively calculate and obtain the global average SoC and average DoD estimates of the energy storage power system. The second calculation unit: Based on the deviation between the SoC of each energy storage battery cell and the global average SoC, and the deviation between the DoD of each energy storage battery cell and the global average DoD, it calculates the result using an error function ( erf The nonlinear formula is used to dynamically calculate the adaptive droop coefficient for each energy storage battery cell. R vi ); Main control layer: Used to generate primary voltage control commands based on the adaptive droop coefficient of each energy storage cell. V dci And through the converter corresponding to each energy storage battery unit, the initial power allocation of each energy storage battery unit and the initial balancing of the SoC are achieved; Secondary control layer: After the initial power allocation and SoC equalization of each energy storage battery cell, a secondary control strategy is adopted. First, the local power allocation influence coefficient of each energy storage battery cell is calculated. λ i ) and information modality factor ( i The system obtains the global average value of the information mode factor through a consensus algorithm; then, it generates a voltage compensation amount based on the bus voltage deviation and the information mode factor deviation to correct the primary voltage control command. The generated final voltage control command is applied to the bidirectional DC-DC converter corresponding to each energy storage battery unit to achieve precise power distribution of the energy storage battery unit and recovery of the bus voltage.

[0025] like Figure 1 The diagram shown is a block diagram of the new energy ship battery power system of the present invention. Figure 1 The demonstration showcased the power architecture of an all-electric vessel, integrating a photovoltaic array, parallel batteries with converters, and propulsion / auxiliary loads via a DC bus. The power system is connected to the DC bus via parallel converters and employs segmented switching equipment for advanced fault isolation. The photovoltaic array is connected to the DC bus via a boost converter, while the BSUs are connected via bidirectional DC / DC converters.

[0026] like Figure 2 The diagram shown is a topology of the hierarchical coordinated control method proposed in this invention. It includes a main control layer, a secondary control layer, and a communication layer. Each battery is connected in parallel with the bus, and the relevant average data calculated from adjacent batteries is transmitted through each BSU (Bus Unit). Figure 2 It is connected to the DC bus via a bidirectional DC-DC converter. Adjacent units communicate via low-bandwidth communication (LBC).

[0027] The implementation of the physical layer includes the following key components: L i Indicates the input-side inductance. C i This represents the output filter capacitor. The switching device is controlled by the gate signal. Q 1 and Q 2 drives, r i Represents the equivalent line impedance. Current measurement includes... i Li andi dci This provides crucial feedback for system control.

[0028] The control layer adopts a hierarchical design. The first-level control layer uses SoC adaptive droop control to achieve dynamic balancing, while the second-level control layer compensates for current distribution errors and regulates bus voltage by coordinating load power distribution and voltage compensation. The communication layer constructs an adjacent communication network structure based on a dynamic consensus algorithm. Each BSU only interacts with its neighboring units to collaboratively estimate global variables and converge to the target value.

[0029] like Figure 3 The diagram shows the communication topology between distributed batteries in this invention. To reduce system communication pressure, each battery is regarded as a multi-agent, and then an adjacent-to-adjacent sparse communication network is constructed. A dynamic consensus algorithm is used to obtain the average information of the system.

[0030] like Figure 4 The image shows the charging mode in this invention. R vi The surface plots show two key operational characteristics of the proposed adaptive droop control: R vi In fixed ε In the case of increasing k It will expand the high SoC unit R vi (Reduce its charging current), while also reducing the lower SoC battery cell's... R vi ; With k fixed, reduce ε It will improve R vi Approaching equilibrium ( SoC i The adjustment rate when ≈0) is used to accelerate the convergence speed. Both of these behaviors are in line with the basic principles of droop control.

[0031] like Figure 5 The figure shows the discharge mode in this invention. R vi Surface plots, analysis methods and Figure 4 similar.

[0032] like Figure 6 The figure shows the discharge droop curve of BSUs in this invention, which fully verifies the effectiveness of the proposed strategy. At the beginning of discharge, due to line impedance mismatch, BSU1 / BSU2 operates at point A1 / A2, resulting in inaccurate output current distribution and bus voltage deviation. The control strategy is adjusted in two stages: i) through δv i1 / δ vi2 ii) Balance the current, flattening the curve to point A; ii) restore the voltage by shifting the curve upward to point B. This scheme achieves simultaneous voltage stabilization and precise current balancing through impedance / voltage coordinated compensation.

[0033] like Figure 7 (a)(b)(c)(d) show the simulation results under propulsion and auxiliary load fluctuations in this invention.

[0034] In this scenario, the initial SoC of the BSUs are 80%, 70%, 60%, and 50%, respectively. Each BSU has a rated capacity of 6Ah.

[0035] Figure 7 The simulation results for propulsion and auxiliary load fluctuations in this invention are shown, where (a) output power, (b) SoC of the lithium battery, (c) output current, and (d) DC bus voltage. The simulation results for Scenario 1 are shown.

[0036] The simulation of Scenario 1 consists of six stages, as detailed in Table 1 below.

[0037] Table 1 Simulation Process of Scenario 1

[0038] from Figure 7 (b)(c)(d) It can be seen that SoC equalization and output current share are... t The DC bus voltage is achieved in approximately 39 seconds, with a maximum deviation from the reference value of about 15V. In the event of a sudden load change in the propulsion or auxiliary systems, the DC bus voltage may momentarily deviate from its reference value, but it will still be able to... V dc It maintains strict regulation within a range of ±10% (equivalent to ±75V).

[0039] Simulation results for Scenario 1 show that, under the condition of consistent BSU capacity, the proposed control strategy achieves fast SoC equalization, accurate current distribution, and stable bus voltage regulation, while keeping the bus voltage within a reasonable range during load fluctuations.

[0040] like Figure 8 The simulation results for SoC equalization speed comparison in this invention are shown, where (a) is the proposed strategy; and (b) are references. [1] Strategy; (c) References [2] Strategy; (d) References [3] Strategy; In this scenario, comparison with the literature confirms that the proposed control strategy achieves faster SoC equalization. The initial SoCs of the BSUs are 80%, 70%, 60%, and 50%, respectively. The rated capacity of all BSUs is 10Ah. The simulation of Scenario 2 consists of three stages, as detailed in Table 2 below.

[0041] Table 2 Simulation Process of Scenario 2

[0042] Figure 8 (a)(b)(c)(d) indicate that the SoC under the proposed control strategy is approximately t Near-complete convergence was achieved at 17s. In subsequent simulations, [the following method was used]. Figure 8 (a) The same method. However, from Figure 8 As can be seen from (b)(c)(d), when using the droop control methods in the comparative references, BSU1 and BSU4 perform better. t The SoC differences at 17s were approximately 10.21%, 9.47%, and 10.72%, respectively. Furthermore, only some of the references reached equilibrium at the assumed end of the time period.

[0043] Simulation results for Scenario 2 show that, under the same operating conditions, the proposed control strategy can achieve better SoC balanced performance, effectively prevent BSUs from overcharging / discharging, and significantly improve cycle life and system stability.

[0044] Figure 9 The simulation results under the plug-and-play test in this invention are shown, where (a) output power; (b) SoC of the lithium battery; (c) output current; and (d) DC bus voltage. This scenario mainly tests the plug-and-play performance of the proposed control strategy. The initial SoCs of the BSUs were 80%, 70%, 60%, and 50%, respectively. The rated capacity of all BSUs was 5Ah. BSU3 in t The simulation will exit after 15-20 seconds. Scenario 3 consists of 8 stages, detailed in Table 3 below.

[0045] Table 3 Simulation Process of Scenario 3

[0046] Figure 9 (a)(b)(c)(d) show the simulation results for scenario three. Figure 9 It can be seen from (b)(c)(d) that, t The simulation process before 15 seconds is similar to Scheme 1. SoC3 in t BSU3 stops changing after 10 seconds. tOperation resumed after 20 seconds, and no significant fluctuations were observed in the DC bus. The analysis results demonstrate that the proposed control strategy possesses excellent plug-and-play performance. Furthermore, when... t At 10s, the propulsion load decreases, and the DC bus voltage deviation is 17V; when t At 25s, the propulsion load increased, the DC bus voltage deviation was 18V, and the overshoot was within the normal range.

[0047] Simulation results for Scenario 3 demonstrate that, under the same BSU capacity conditions, the proposed control strategy can simultaneously achieve SoC equalization and precise current distribution, effectively compensate for line impedance effects, and maintain stable bus voltage. The system exhibits robust operation during BSU failures / reconnections, proving its excellent scalability.

[0048] Figure 10 The simulation results under photovoltaic fluctuations and communication anomalies in this invention are shown, where (a) output power, (b) lithium battery SoC, (c) output current, and (d) DC bus voltage. The scheme tested photovoltaic fluctuations, communication anomalies, and BSU capacity, respectively. C b1 =5Ah C b2 =10Ah C b3 =15Ah and C b4 The control effect at 20Ah (initial SoC is 50%). The simulation of Scenario 4 consists of three stages, as detailed in Table 4 below.

[0049] Figure 10 (b)(c)(d) demonstrate that traditional droop control cannot maintain SoC balance; under unbalanced operating conditions, line impedance mismatch leads to gradual divergence. The proposed control strategy ensures asymptotic SoC convergence while precisely maintaining current sharing among distributed energy storage units according to capacity ratio. Although in t A communication failure occurred between BSU2 and BSU3 at 30 seconds, but the system still maintained stable operation. t At 36 seconds, the SoC recovered equilibrium, proving that the control strategy has a certain fault tolerance capability.

[0050] Table 4 Simulation Process of Scenario 4

[0051] Simulation results for Scenario 4 show that, under different BSU capacity and irradiance conditions, the proposed control strategy can maintain SoC convergence and bus voltage regulation, while the dynamic consensus algorithm can maintain normal system operation even in the event of communication failure.

[0052] Figure 11 The simulation results for the ferry navigation scenario in this invention are shown, where (a) output power, (b) SoC of the lithium battery, (c) output current, and (d) DC bus voltage. This study verified the feasibility of the proposed control strategy by simulating actual ferry operating conditions. The BSUs were initialized with initial SoC levels (90%, 80%, 70%, and 60%) to evaluate performance under conditions of inconsistent actual SoC. All BSUs had a rated capacity of 10 Ah.

[0053] The simulation of Scenario 5 consists of six stages, as detailed in Table 5 below.

[0054] Table 5 Simulation Process of Scenario 5

[0055] Figure 11 (a)(b)(c)(d) show the simulation results for the fifth scenario. t Achieve SoC equalization and precise distribution of output current within 12 seconds.

[0056] Simulation data from Scenario 5 show that, among BSUs of the same capacity, the proposed control strategy achieves fast SoC equalization, accurate current distribution, and reliable bus voltage recovery, meeting the operational requirements of the assumed ferry propulsion scenario.

[0057] Scenario 6 is a scalability test for all-electric propulsion ships; the simulation of Scenario 6 consists of three stages, as detailed in Table 6 below.

[0058] Table 6 Simulation Process of Scenario 6 This scenario evaluates the scalability of the proposed control strategy using a system of twelve battery packs. Each battery unit (BSU) has a rated capacity of 5Ah and initial (SoC) values ​​of 88%, 85%, 82%, 76%, 73%, 70%, 67%, 64%, 61%, 57%, 54%, and 51%, respectively.

[0059] Figure 12 The simulation results for the all-electric propulsion ship battery scalability scenario in this invention are shown, where (a) output power, (b) lithium battery SoC, (c) output current, and (d) DC bus voltage. The proposed control strategy achieves both charge balance and precise output current distribution simultaneously within approximately 43 seconds, with a maximum bus voltage deviation of 14V, meeting the overshoot requirement. The results confirm that this strategy can achieve rapid balancing and precise current distribution, demonstrating its applicability to engineering applications and its significant advantages in systems with a large number of batteries.

[0060] Figure 13 The experimental results under propulsion and auxiliary load fluctuations in this invention are shown, wherein (a) the SoC of the lithium battery; (b) the output current; and (c) the DC bus voltage. In Scenario 1 ( C b1 = C b2 = C b3 = C b4 Under experimental conditions matching 12Ah, Figure 13 (a)(b)(c) demonstrate the proposed control strategy's SoC equalization, precise current distribution, and stable voltage recovery. The experimental results are in good agreement with the simulation predictions.

[0061] Figure 14 The experimental results for different energy storage capacities in this invention are shown, where (a) the SoC of the lithium battery; (b) the output current; and (c) the DC bus voltage. Under experimental conditions matching Scenario 4, Figure 14 (a)(b)(c) demonstrate the performance of the proposed control strategy in handling photovoltaic fluctuations, communication anomalies, and different capacities of BSUs. The experimental results are in good agreement with the simulation data of Scenario 4.

[0062] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

[0063] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

[0064] [1] Q. J. Zhang et al., "An Improved Distributed Cooperative ControlStrategy for Multiple Energy Storages Parallel in Islanded DC Microgrid,"IEEE JOURNAL OF EMERGING AND SELECTED TOPICS IN POWER ELECTRONICS, vol. 10,no. 1, pp. 455-468, FEB 2022. [2] R. Iqbal, Y. C. Liu, Y. J. Zeng, Q. J. Zhang, and H. Y. Yu, "Adaptive droop-based SoC balancing control scheme for parallel batterystorage system in shipboard DC micro grid," JOURNAL OF ENERGY STORAGE, vol.79, FEB 15 2024, Art. no. 110205. [3] T. R. Oliveira, W. W. A. G. Silva, and P. F. Donoso-Garcia, "Distributed Secondary Level Control for Energy Storage Management in DC Microgrids," IEEE Transactions on Smart Grid, vol. 8, no. 6, pp. 2597-2607, 2017.

Claims

1. A hierarchical coordinated control method for the state-of-charge balance of parallel energy storage batteries in an all-electric propulsion ship, characterized in that, Includes the following steps: S1: Obtain the parameters of multiple energy storage battery units and their corresponding connected bidirectional DC-DC converters, and establish a distributed communication network based on the consensus of energy storage battery units; S2: Obtain the initial SoC, capacitance, charge / discharge efficiency, and current of the energy storage battery cell in the distributed communication network, and use the coulomb integral method to calculate the state of charge and depth of discharge of the energy storage battery cell. S3: Based on energy storage battery cells, the state of charge and depth of discharge of adjacent energy storage battery cells are obtained through low-bandwidth communication links. The global average state of charge and average depth of discharge of the energy storage power system are obtained by iterative calculation using a dynamic consensus algorithm. S4: Based on the deviation between the state of charge of each energy storage battery cell and the global average state of charge, and the deviation between the depth of discharge of each energy storage battery cell and the global average depth of discharge, the adaptive droop coefficient of each energy storage battery cell is dynamically calculated using a nonlinear formula that includes an error function. S5: Based on the adaptive droop coefficient of each energy storage battery cell, a primary voltage control command is generated and executed by the converter corresponding to each energy storage battery cell to achieve the initial power allocation of each energy storage battery cell and the initial SoC balance. S6: After the initial power allocation and initial balancing of each energy storage battery cell, based on the secondary control strategy, the local power allocation influence coefficient and information mode factor of each energy storage battery cell are first calculated, and the average value of the information mode factor is obtained through the consensus algorithm. Then, the voltage compensation amount is generated by integrating the quotient of the average value of the information mode factor and the power allocation influence coefficient minus the bus voltage reference value. Finally, the primary voltage control command is corrected to generate the final voltage control command, thereby realizing the accurate power allocation of the energy storage battery cell and the recovery of the bus voltage.

2. The method for hierarchical coordinated control of the state of charge balance of parallel energy storage batteries in an all-electric propulsion ship according to claim 1, characterized in that, The expression for calculating the state of charge of the energy storage battery cell using the Coulomb integral method is as follows; (1) In equation (1), BSU i The initial SoC, BSU i The charge and discharge efficiency, BSU i The charging and discharging current, BSU i The capacity.

3. The method for hierarchical coordinated control of the state of charge balance of parallel energy storage batteries in an all-electric propulsion ship according to claim 1, characterized in that, The expression for the dynamic consensus algorithm is: (2) In formula (2) X i =[SoC i , φ i ] indicates BSU i The calculated data, N i It is the set of its neighboring nodes. a ij It is a communication weighting coefficient; X avg _ i ( k )and X avg _ i ( k +1) represent iterations. k and k Local average estimate at +1; BSU i and BSU j The cumulative error value between them is used D ij ( k ) indicates that it is initialized to D ij (0) = 0, while binary variables δ ij ∈{0,1} indicates whether there is a communication link between the corresponding nodes.

4. The method for hierarchical coordinated control of the state of charge balance of parallel energy storage batteries in an all-electric propulsion ship according to claim 1, characterized in that, When the current of the energy storage battery cell is less than 0, the droop coefficient is calculated using the following formula: (3) In equation (3), C bati , max BSU i Maximum capacity, C bati BSU i Rated capacity, R v0 This is the initial droop coefficient. erf [ k ρ SoC i ]、 erf [ k ρ DoD i ] is the error function. k For convergence speed, ρ For acceleration factor, SoC i It's BSU i SoC, DoD i It's BSU i Discharge depth, ΔDoD i It is the difference between the depth of discharge and the average DoD, SoC a and DoD a They are BSU i The average state of charge and average depth of discharge.

5. The method for hierarchical coordinated control of the state of charge balance of parallel energy storage batteries in an all-electric propulsion ship according to claim 1, characterized in that, When the current of the energy storage battery cell is ≥0, the formula for calculating the droop coefficient is: (4) In equations (3) and (4), (5) (6) In equation (6), ε For convergence accuracy; (7) (8) (9)。 6. The method for hierarchical coordinated control of the state of charge balance of parallel energy storage batteries in an all-electric propulsion ship according to claim 1, characterized in that, Generate primary voltage control instructions V dci ,for: (10) Utilizing generated voltage compensation amount δv i The primary voltage control command is modified to simultaneously achieve precise power distribution and bus voltage recovery and stabilization. The modified primary voltage control command is expressed as follows: (11)。 7. The method for hierarchical coordinated control of the state of charge balance of parallel energy storage batteries in an all-electric propulsion ship according to claim 1, characterized in that, The secondary control strategy employs a single integral controller to simultaneously ensure accurate current distribution and bus voltage recovery. The specific method is as follows: (12) In equation (12), V dc The bus voltage reference value is set to 750V. avg for i The average value; λ i For the first i The power distribution influence coefficient of the converter, where avg The expression is shown below. (13) In equation (13), i It is the first i The information mode factor of each converter, and i = λ i × V dci ,in λ i The expression is shown below. (14) In formula (14) P pui The expression is as follows: (15) In equations (14) and (15), P pui For unit output power, R vi For adaptive droop coefficient, b It is a constant. R vi_max It is the maximum value of the adaptive droop coefficient under the set relationship curve between balancing accuracy and balancing speed. P dci BSU i 'output power' P ratei BSU i Rated output power, b It is a constant, and its function is to prevent when R vi P dci = R vi_max P ratei When the denominator is 0, R vi P dci / R vi_max P ratei express R vi P dci and R vi_max P ratei The proportion is to i Limited to the range of 0-1, and must satisfy 1- R vi P dci / b R vi_max. .

8. A new energy ship, characterized in that, According to any one of claims 1-7, the hierarchical coordinated control method for the state-of-charge balance of parallel energy storage batteries in all-electric propulsion ships enables the control of the power system.

9. A hierarchical coordinated control device for the state-of-charge balance of parallel energy storage batteries in a fully electric propulsion ship. Its characteristics include: The first acquisition module acquires parameters of multiple energy storage battery units and their corresponding connected bidirectional DC-DC converters, and establishes a distributed communication network based on the consensus of the energy storage battery units. The second acquisition module acquires the initial state of charge, capacitance, charge / discharge efficiency, and current of the energy storage battery cells in the distributed communication network, and uses the coulomb integral method to calculate the state of charge and depth of discharge of the energy storage battery cells. The first calculation unit: Based on the energy storage battery unit (BSU), it exchanges state of charge and depth of discharge information with adjacent units through a low-bandwidth communication link. Then, based on the state of charge and depth of discharge of all energy storage battery units, it uses a dynamic consensus algorithm to iteratively calculate and obtain the global average state of charge and average DoD estimate of the energy storage power system. The second calculation unit dynamically calculates the adaptive droop coefficient of each energy storage battery cell based on the deviation between the state of charge of each energy storage battery cell and the global average state of charge, and the deviation between the depth of discharge of each energy storage battery cell and the global average depth of discharge, using a nonlinear formula that includes an error function. Main control layer: Used to generate primary voltage control commands based on the adaptive droop coefficient of each energy storage battery cell, and execute them through the corresponding converter of each energy storage battery cell to achieve the initial power allocation of each energy storage battery cell and the initial SoC balancing; Secondary control layer: After the initial power allocation and initial balancing of each energy storage battery cell, based on the secondary control strategy, the local power allocation influence coefficient and information mode factor of each energy storage battery cell are first calculated, and the average value of the information mode factor is obtained through the consensus algorithm. Then, the voltage compensation amount is generated by integrating the quotient of the average value of the information mode factor and the power allocation influence coefficient minus the bus voltage reference value. Finally, the primary voltage control command is corrected to generate the final voltage control command, so as to realize the accurate power allocation of the energy storage battery cell and the recovery of the bus voltage.

Images