Ammonia-hydrogen fuel cell hybrid energy storage control method for optimizing operation cost and island stability

Abstract

The invention provides an ammonia-hydrogen fuel cell hybrid energy storage control method for optimizing operation cost and island stability, and the method comprises the steps: solving the power distribution which meets the constraint conditions of a hybrid energy storage system and has the lowest operation cost in real time based on the island fuel cell-lithium battery hybrid energy storage system; and a self-adaptive VSG control strategy which simultaneously considers energy storage side response delay response capability and lithium battery pack SOC limitation is adopted to carry out network construction type self-adaptive VSG control on an inverter in the isolated island fuel cell-lithium battery hybrid energy storage system. According to the invention, the hybrid energy storage system based on the ammonia-hydrogen fuel cell and the lithium battery is applied to the island micro-grid system, so that the comprehensive goals of long endurance, high power, high efficiency and low cost are achieved; frequency and voltage support is provided for the island system through network construction type VSG control, and independent power supply for the load can be realized without depending on an external power distribution network; meanwhile, the operation cost of the system is effectively reduced, and frequent fuel supply is avoided.

Description

Technical Field

[0001] This invention relates to the field of fuel cell control, and more specifically to a hybrid energy storage control method for ammonia-hydrogen fuel cells that optimizes operating costs and islanding stability. Background Technology

[0002] Existing energy management systems typically focus only on energy allocation and cost optimization during steady-state operation, neglecting transient stability issues. While feasible in grid-connected energy storage systems, this approach is unsuitable for islanded energy storage systems. VSG control strategies, commonly used inverter control strategies in islanded microgrids, have evolved into adaptive VSG control strategies capable of effectively handling various situations. However, most existing adaptive VSG control strategies adaptively adjust the VSG's inertia and damping coefficients by sampling and measuring the inverter's output voltage and current in real-time using sensors, calculating the current rate of change of angular frequency, deviation, and active power difference. These are trigger-based adaptive adjustment methods based on frequency event influences, often accompanied by millisecond-level time delays in sampling, communication, and calculation in practical engineering. This is especially problematic when power fluctuates rapidly and significantly, as metrics such as the rate of change of angular frequency and active power difference cannot instantly reflect the system's current state. This is particularly evident in islanded VSG systems. Unlike grid-connected VSG systems, all inertia and damping in islanded VSG systems are provided by the VSG itself, and existing adaptive VSG control methods cannot provide rapid support on a millisecond-level timescale. Meanwhile, most existing adaptive VSG control methods assume that the instantaneous power response capability of the energy storage side can fully cover the power changes under the target scenario. This application considers the problem caused by the energy storage response delay when the instantaneous power response of the energy storage side is insufficient.

[0003] Existing energy management strategies for fuel cell hybrid energy storage systems mostly consider fuel consumption costs and system maintenance costs, but the system maintenance costs are not adequately detailed. This application refines this part of the maintenance cost, mainly consisting of the degradation costs of each energy storage unit, and further refines the degradation calculation of the fuel cell, which is beneficial for the system to further seek the optimal solution. Summary of the Invention

[0004] The purpose of this invention is to propose a hybrid energy storage control method for ammonia-hydrogen fuel cells that optimizes operating costs and islanding stability. This method applies a hybrid energy storage system based on ammonia-hydrogen fuel cells and lithium batteries to islanded microgrid systems, achieving a comprehensive goal of long range, high power, high efficiency, and low cost. By using grid-type VSG control, it provides frequency and voltage support to the islanded system, enabling independent power supply to loads without relying on an external distribution network. Simultaneously, it effectively reduces system operating costs and avoids frequent refueling.

[0005] To achieve the above objectives, the technical solution of the present invention is: a hybrid energy storage control method for ammonia-hydrogen fuel cells that optimizes operating costs and islanding stability. Based on an islanded fuel cell-lithium battery hybrid energy storage system, the method solves in real time the power allocation that minimizes operating costs while satisfying the constraints of the hybrid energy storage system. An adaptive VSG control strategy that simultaneously considers the response delay capability of the energy storage side and the SOC limit of the lithium battery pack is adopted to perform grid-type adaptive VSG control on the inverter in the islanded fuel cell-lithium battery hybrid energy storage system.

[0006] Preferably, the islanded fuel cell-lithium battery hybrid energy storage system includes an ammonia-hydrogen fuel cell, a power control module, a lithium battery pack, a voltage control module, an inverter, an adaptive VSG control system, an energy management system, and a load;

[0007] The ammonia-hydrogen fuel cell is connected to the inverter via a Boost converter and employs power control.

[0008] The lithium battery pack is connected to the inverter via a bidirectional DC-DC module and is controlled by voltage.

[0009] The energy management system calculates the optimal power allocation based on the input data, system constraints, and system operating costs, and outputs the fuel cell reference power P. fc The power control module; the input data of the energy management system includes the inverter output power P. e Real-time SOC of lithium battery pack, output voltage of lithium battery pack (V) ba Fuel cell output voltage V fc Fuel cell output power P' fc ;

[0010] Inverter output power P e The calculation is as follows:

[0011] P e =P m -ω·J dω / dt-ω·D·(ω-ω0)

[0012] Where: P m Let J be the mechanical power of VSG, and D be the virtual inertia and damping coefficient, respectively. ω and ω0 be the virtual angular frequency and the rated angular frequency, respectively.

[0013] The power control module is based on the input fuel cell reference power P. fc Output PWM signal for power control of ammonia-hydrogen fuel cell;

[0014] The voltage control module outputs a PWM signal based on the input DC bus voltage for lithium battery pack voltage control; the DC bus voltage refers to the voltage value of the common DC link connecting the output terminal of the bidirectional DC-DC module and the DC input terminal of the inverter.

[0015] The inverter is controlled by an adaptive VSG control system, and the output is filtered by an LC filter before supplying power to the load.

[0016] The adaptive VSG control system includes an adaptive J and D adjustment module, which is used to adjust the coefficients J and D according to the energy storage status, load demand, and AC side stability.

[0017] Preferably, the quadratic sequential programming algorithm SQP is used to solve for the power allocation with the lowest operating cost that satisfies the constraints of the hybrid energy storage system at each time interval dt;

[0018] System constraints include power balance equation constraints, fuel cell current change rate constraints, lithium battery pack current change rate constraints, fuel cell power upper limit constraints, and lithium battery pack power upper and lower limit constraints.

[0019] The system operating cost includes fuel consumption cost, fuel cell degradation cost, lithium battery pack degradation cost, lithium battery pack SOC maintenance cost, lithium battery pack SOC gradient limit penalty cost, and lithium-ion battery pack adjustable power constraint cost. The adjustable power constraint cost of the lithium-ion battery pack is inversely proportional to the adjustable power of the lithium-ion battery pack.

[0020] Preferably, the specific calculations for fuel consumption cost, fuel cell degradation cost, lithium battery pack degradation cost, and lithium battery pack SOC maintenance cost are as follows:

[0021] Fuel consumption cost

[0022]

[0023] Where: λ H2 This represents the fuel consumption cost coefficient; m H2 The hydrogen fuel consumption cost of the system in the current time period dt; The market price of hydrogen fuel; N is the number of individual cells in the fuel cell; F is the Faraday constant; I fc (t) represents the fuel cell output current, I fc =P′ fc / V fc ;

[0024] Fuel cell degradation costs

[0025]

[0026] Where: λ fc For fuel cell degradation cost coefficient; SOD b (t) is the degradation formula for the fuel cell, representing the degree of degradation of the fuel cell at the current fuel cell power level; c0, c1, and c2 are the coefficients of the quadratic function; P′ fc (t) represents the fuel cell output power at time t; x fc This refers to the output voltage of the fuel cell;

[0027] Degradation cost of lithium battery packs

[0028]

[0029] Where: λ ba ΔQ is the degradation cost coefficient for lithium battery packs. loss Let be the degree of degradation of the lithium battery pack within the current time period dt, i.e., the cycle aging of the battery; E is ; a, b, c, d′, e′ are empirical coefficients; T e Absolute temperature; Ah t0 For throughput per hour; I rate I represents the charge / discharge rate (expressed as C-rate, which is the ratio of charge / discharge current to rated capacity). rate =P ba / (V ba ·Q ba ), where Q ba P represents the rated capacity of the lithium battery pack. ba =P e -P′ fc P e This refers to the inverter's output power.

[0030] Lithium-ion battery pack SOC maintenance cost:

[0031]

[0032] Where: λ SOC Maintain the cost factor for the state of charge (SOC) of the lithium battery pack.

[0033] Preferably, the adaptive VSG control specifically involves executing the following steps in cycles with a time interval dt:

[0034] Step S1: Start the round and start timer t1;

[0035] Step S2: Parameter initialization;

[0036] Step S3: Obtain the adjustable power ΔP of the current lithium battery pack. Ba ↑↓, Real-time SOC of lithium battery pack, Load power P load Inverter output power Pe The angular frequency ω of the VSG is measured.

[0037] Step S4: Determine whether the adjustable power of the current lithium battery pack meets the system power response deficit based on the energy storage response lag criterion; if it meets the energy storage response lag criterion, start timer t2 and proceed to step S5; otherwise, proceed to step S7.

[0038] Step S5: Determine the duration τ of the inertial response phase in the VSG frequency event based on the power disturbance level;

[0039] Step S6: Calculate and output the virtual inertia J and damping coefficient D in real time until t1>τ, then execute step S7; the calculation of virtual inertia J and damping coefficient D is as follows:

[0040] J = J0 + J ΔP +J soc

[0041] D = D0 + D ΔP

[0042] Where: J0 is the reference inertia coefficient; D0 is the reference damping coefficient; J ΔP and D ΔP The adaptive J and D components are characterized by the adjustable power of the lithium battery pack; J soc This considers the J component of battery SOC in system transient events;

[0043] Step S7: Obtain the deviation ΔP of the VSG output active power, the difference Δω between the virtual angular frequency ω and the rated angular frequency ω0, and the rate of change of angular frequency dω / dt; calculate and output the virtual inertia J and damping coefficient D using the following expressions:

[0044] J = J1 + J SOC

[0045] D = D1

[0046] Where: J1 and D1 are the adaptive J and D coefficients without considering SOC;

[0047] Step S8: Determine if t1>T is true. If true, the round ends, where T is the round duration, T = dt; otherwise, update the real-time SOC of the lithium battery pack and the inverter output power P. e And measure the angular frequency ω of the VSG and return to step S7.

[0048] Preferably, the energy storage response hysteresis criterion is:

[0049] ΔP Ba ↑↓≤ρΔP ref

[0050]

[0051] P ref =P load

[0052] Where: ρ is the adjustment coefficient; ΔP ref The reference power change rate of the VSG; These are the reference power of the VSG in the current round and the reference power of the VSG in the previous round, respectively; the reference power P of the VSG. ref The value is equal to the load power P load .

[0053] Preferably, the reference power change rate ΔP of the VSG is used. ref Determine the power disturbance level and assess the duration τ of the inertial response phase in a VSG frequency event:

[0054] If ΔP ref If the value is less than 10%, it is considered a small disturbance, and the duration τ of the inertial response phase in the VSG frequency event is set to τ1.

[0055] If 10% ≤ ΔP ref If the value is ≤30%, it is considered a moderate disturbance, and the duration τ of the inertial response phase in the VSG frequency event is set to τ2.

[0056] If ΔP ref If the value is greater than 30%, it is considered a large disturbance, and the duration τ of the inertial response phase in the VSG frequency event is set to τ3.

[0057] Among them, τ1<τ2<τ3.

[0058] Preferably, the calculation of the adaptive J and D components, characterized by the adjustable power of the lithium battery pack, is as follows:

[0059]

[0060] Where: F is the power response coefficient of the fuel cell; θ1 are the adjustment coefficients for the J and D components, respectively.

[0061] Preferably, the calculation of the J component of battery SOC during system transient events is considered, as follows:

[0062]

[0063] in: The adjustment coefficient; SOC ref This is a reference SOC for lithium battery packs.

[0064] Preferably, the calculation of the adaptive J and D coefficients of SOC is not considered, as follows:

[0065]

[0066] Where: K0 is the threshold value for the change of the difference Δω between the virtual angular frequency and the rated angular frequency during J parameter adjustment; K1 is the threshold value for the change of dω / dt during J parameter adjustment; K D Δω represents the threshold value for the change in D parameter during adjustment; k1 to k5 are adjustment coefficients; h is the conversion coefficient between the rate of change of angular frequency and the power deviation.

[0067] Compared with the prior art, the present invention has the following beneficial effects:

[0068] The criterion for instantaneous power response capability of fuel cell hybrid energy storage system proposed in this invention can achieve adaptive control earlier than the active power-frequency event in islanded VSG system, while avoiding the delay caused by system sampling and calculation, thus achieving the most timely adaptive adjustment.

[0069] The energy management system of this invention not only considers the cost of fuel consumption, but also the degradation cost of energy storage devices. The calculation of the degradation cost of fuel cells is different from the common method based on degradation mechanism partitioning. It fits the degradation curve of fuel cells, making the calculation of its degradation cost more accurate. Attached Figure Description

[0070] Figure 1 This is a framework diagram of the islanded fuel cell-lithium battery hybrid energy storage system of the present invention;

[0071] Figure 2 This is a block diagram of the fuel cell power control of the present invention;

[0072] Figure 3 This invention relates to the DC bus voltage control of the lithium battery pack.

[0073] Figure 4 This is a block diagram of the adaptive VSG control system of the present invention;

[0074] Figure 5 This is a diagram illustrating the energy management strategy of the present invention.

[0075] Figure 6 The graph shows the changes in the angular frequency and active power of the virtual synchronous generator before and after optimization in this invention.

[0076] Figure 7 This invention provides a comparison of adaptive VSG control in the initial stage of frequency events.

[0077] Figure 8 This is a flowchart of the adaptive VSG control of the present invention. Detailed Implementation

[0078] The following is in conjunction with the appendix Figure 1-8The technical solution of the present invention will be described in detail below.

[0079] Ammonia-hydrogen fuel cells, as a clean and efficient power generation technology, have gradually developed into a high-quality energy storage unit due to their high cost-effectiveness and superior safety compared to traditional fuel cells. However, fuel cells are limited by their inherently low response rate and slow power response, making it difficult to quickly cope with instantaneous high loads. Lithium-ion batteries, with their high power density (capable of rapidly releasing large currents), can provide peak power in a short time, compensating for the response lag of fuel cells. This invention combines ammonia-hydrogen fuel cells with lithium-ion batteries to achieve synergy between "energy-type power sources" and "power-type power sources." Through technological complementarity, it achieves the comprehensive goals of "long range, high power, high efficiency, and low cost," making it particularly suitable for complex scenarios with high requirements for energy density, power response, and reliability. This hybrid system not only breaks through the performance bottlenecks of single-cell technologies but also provides a better solution for the large-scale application of new energy technologies. The hybrid energy storage system based on ammonia-hydrogen fuel cells and lithium-ion batteries is applied to an islanded microgrid system. The inverter control adopts grid-type VSG control, which can provide frequency and voltage support for the islanded system, enabling independent power supply to the load without relying on an external distribution network. At the same time, considering the single nature of the energy storage units in isolated energy storage systems, it is necessary to reduce the operating costs of the system and avoid frequent refueling.

[0080] This invention proposes a hybrid energy storage control method for ammonia-hydrogen fuel cells that optimizes operating costs and islanding stability. Based on an islanded fuel cell-lithium battery hybrid energy storage system, the method solves in real time the power allocation that minimizes operating costs while satisfying the constraints of the hybrid energy storage system. An adaptive VSG control strategy that simultaneously considers the response delay capability of the energy storage side and the SOC limit of the lithium battery pack is adopted to perform grid-type adaptive VSG control on the inverter in the islanded fuel cell-lithium battery hybrid energy storage system.

[0081] like Figure 1 As shown, in this embodiment, the islanded fuel cell-lithium battery hybrid energy storage system includes an ammonia-hydrogen fuel cell, a power control module, a lithium battery pack, a voltage control module, an inverter, an adaptive VSG control system, an energy management system, and a load.

[0082] The ammonia-hydrogen fuel cell is connected to the inverter via a Boost converter, and employs, for example... Figure 2 The power control shown;

[0083] The lithium battery pack is connected to the inverter via a bidirectional DC-DC module, and uses, for example... Figure 3 The voltage control shown; the lithium battery pack stabilizes the DC bus voltage through a bidirectional DC-DC module while simultaneously handling the remaining power demand, i.e., P. e -P' fc, where P load For load power, P' fc This refers to the output power of the fuel cell;

[0084] The energy management system calculates the optimal power allocation based on the input data, system constraints, and system operating costs, and outputs the fuel cell reference power P. fc (Command power) is sent to the power control module; the input data of the energy management system includes the inverter output power P. e Real-time SOC of lithium battery pack, output voltage of lithium battery pack (V) ba Fuel cell output voltage V fc Fuel cell output power P' fc ;

[0085] Inverter output power P e The calculation is as follows:

[0086] P e =P m -ω·J dω / dt-ω·D·(ω-ω0)

[0087] Where: P m Let P be the mechanical power of the VSG. m Numerically equal to the reference power P of the VSG ref The reference power P of the VSG ref Numerically equal to the load power P load P e The latter part of the expression, ω·J dω / dt-ω·D·(ω-ω0), represents the frequency modulation power of the VSG control, where J and D are the virtual inertia and damping coefficient, respectively, and ω and ω0 are the virtual angular frequency and the rated angular frequency, respectively.

[0088] The power control module is based on the input fuel cell reference power P. fc Output PWM signal for power control of ammonia-hydrogen fuel cell;

[0089] The voltage control module outputs a PWM signal based on the input DC bus voltage for lithium battery pack voltage control; the DC bus voltage refers to the voltage value of the common DC link connecting the output terminal of the bidirectional DC-DC module and the DC input terminal of the inverter.

[0090] The inverter is controlled by an adaptive VSG control system, and the output is filtered by an LC filter before supplying power to the load. Figure 1 in,i labc and i gabc These are the inverter output current and the grid connection point current, u abc It is the grid connection point voltage, i labc igabc u abc The five quantities, namely virtual inertia J and damping coefficient D, are processed by the power calculation & VSG module to obtain the input of the voltage and current dual closed-loop control module, and then processed by the PWM modulation stage to obtain the control signal of the inverter DC / AC.

[0091] The control block diagram of the adaptive VSG control system is as follows: Figure 4 As shown, it includes active-frequency control, reactive-voltage control and voltage-current dual closed-loop control, virtual impedance, and adaptive J and D regulation modules;

[0092] The adaptive J and D adjustment module is used to adjust the coefficients J and D according to the energy storage state, load demand, and AC side stability. Specifically, the inputs of the adaptive J and D adjustment module include the reference power change rate ΔP of the VSG. ref Real-time SOC of lithium battery packs and some state variables of energy storage.

[0093] During steady-state operation, the energy management system achieves power allocation with optimal operating cost while meeting basic constraints. When there are significant power fluctuations in load demand, the system considers the energy storage response characteristics to avoid system stability issues caused by the short-term lag of output power behind the reference power. The adaptive VSG control method proposed in this invention is adopted to effectively avoid the frequency stability problem of islanded microgrid systems.

[0094] In actual operation, load demand is monitored and used as the active power reference value for VSG control. The inverter's output demand power is used as the input to the energy management system, which then allocates power in real time. When the system transient event does not trigger the conditions for using the method proposed in this application, the conventional adaptive VSG control method will be adopted. If the conditions for using the method proposed in this application are triggered, the system control will combine the current state of the energy storage system and the changes in load demand to adopt the proposed adaptive VSG control method. At the same time, the energy management system will also incorporate some state variables of the energy storage system into the optimization to minimize the occurrence of condition triggering, that is, to minimize the occurrence of power lag in the energy storage system, thus achieving deep integration of energy management strategy and inverter control.

[0095] In this embodiment, reference Figure 5 The quadratic sequential programming algorithm SQP is used to solve for the power allocation with the lowest operating cost that satisfies the constraints of the hybrid energy storage system at each time interval dt.

[0096] System constraints include power balance equation constraints, fuel cell current change rate constraints, lithium battery pack current change rate constraints, fuel cell power upper limit constraints, and lithium battery pack power upper and lower limit constraints.

[0097] The system operating cost includes fuel consumption cost, fuel cell degradation cost, lithium battery pack degradation cost, lithium battery pack SOC maintenance cost, lithium battery pack SOC gradient limit penalty cost, and lithium-ion battery pack adjustable power constraint cost. The adjustable power constraint cost of the lithium-ion battery pack is inversely proportional to the adjustable power of the lithium-ion battery pack.

[0098] The adjustable power constraint cost of a lithium-ion battery pack represents the equivalent cost corresponding to the current adjustable power of the lithium-ion battery pack. For example, if the current adjustable power of the lithium-ion battery pack is very small, the system may not be able to handle the subsequent load changes instantly. The smaller the adjustable power, the higher the cost. This will prompt the system to increase the adjustable power as much as possible to reduce the system cost and better cope with possible load changes. In other words, the adjustable power constraint cost of a lithium-ion battery pack is inversely proportional to the adjustable power of the lithium-ion battery pack.

[0099] In this embodiment, the specific calculations for fuel consumption cost, fuel cell degradation cost, lithium battery pack degradation cost, and lithium battery pack SOC maintenance cost are as follows:

[0100] Fuel consumption cost

[0101]

[0102] Where: λ H2 This represents the fuel consumption cost coefficient; m H2 The hydrogen fuel consumption cost of the system in the current time period dt; The market price of hydrogen fuel; N is the number of individual cells in the fuel cell; F is the Faraday constant; I fc (t) represents the fuel cell output current, I fc =P′ fc / V fc ;

[0103] Fuel cell degradation costs

[0104]

[0105] Where: λ fc For fuel cell degradation cost coefficient; SOD b (t) is the degradation formula for the fuel cell, representing the degree of degradation of the fuel cell at the current fuel cell power level; c0, c1, and c2 are the coefficients of the quadratic function; P′ fc (t) represents the fuel cell output power at time t; x fc This refers to the output voltage of the fuel cell;

[0106] Degradation cost of lithium battery packs

[0107]

[0108] Where: λ ba ΔQ is the degradation cost coefficient for lithium battery packs. loss Let be the degree of degradation of the lithium battery pack within the current time period dt, i.e., the cycle aging of the battery; E is ; a, b, c, d′, e′ are empirical coefficients; T e Absolute temperature; Ah t0 For throughput per hour; I rate I represents the charge / discharge rate (expressed as C-rate, which is the ratio of charge / discharge current to rated capacity). rate =P ba / (V ba ·Q ba ), where Q ba P represents the rated capacity of the lithium battery pack. ba =P e -P′ fc P e This refers to the inverter's output power.

[0109] Lithium-ion battery pack SOC maintenance cost:

[0110]

[0111] Where: λ SOC Maintain the cost factor for the state of charge (SOC) of the lithium battery pack.

[0112] The adaptive VSG strategy proposed in this invention is based on the conventional adaptive VSG control approach, and simultaneously considers the response delay capability of the energy storage side and the SOC limitation of the lithium battery pack. The lithium battery pack, as a fast-response energy storage unit to compensate for the power response lag of the fuel cell, typically has a lower rated power and capacity than the fuel cell. When the load demand power changes drastically, limited by the adjustable power range of the lithium battery pack at the current moment, part of the response power will be handled by the fuel cell. If the power response from the fuel cell is large, the response capability of the energy storage system is considered insufficient, leading to a situation where the actual output power of the inverter lags behind the reference power, resulting in frequency fluctuations and system instability. Figure 6 To optimize the angular frequency and active power variation curves of the virtual synchronous generator before and after, P0 represents the active power output of the inverter at the previous moment, and P1 represents the current VSG active power reference value. The adaptive VSG control strategy in this invention primarily optimizes the fluctuations of the frequency event in stage 1 (the inertial response stage of the VSG frequency event), suppressing the rate of frequency change and offset in stage 1, through supplementary... Figure 6 It can be seen that the angular frequency and active power curves before optimization are oscillating, while after optimization they converge quickly to a steady state.

[0113] The active-frequency equation for VSG is as follows:

[0114]

[0115] Δω is the difference between the virtual angular frequency and the rated angular frequency; J is the virtual inertia; D is the damping coefficient. Frequency events in islanded VSG systems are mainly due to P m With P e The deviation causes a faster and larger frequency change rate in the initial stage of a frequency event. Let ΔP = P ref -P e Δω=ω-ω0, the above equation can be rearranged as:

[0116]

[0117] It can be seen that in stage 1, which is the first stage of the disturbance, appropriately increasing the J and D parameters is beneficial to reducing the frequency change rate and the frequency change amount. Figure 7 A comparison of adaptive VSG with other adaptive VSG methods that consider energy storage response capability in the early stages of frequency events reveals that conventional adaptive VSG control is based on monitoring the impact before adaptive adjustment. In contrast, the method proposed in this invention is an adjustment method based on load changes and the current rapid response capability of the energy storage side. This method allows adaptive adjustment and frequency events to act simultaneously or even in advance in engineering practice, making it a pre-adjustment method.

[0118] In this embodiment, reference Figure 8 The adaptive VSG control specifically involves executing the following steps in cycles with a time interval dt:

[0119] Step S1: Start the round and start timer t1;

[0120] Step S2: Parameter initialization;

[0121] Step S3: Obtain the adjustable power ΔP of the current lithium battery pack. Ba ↑↓, Real-time SOC of lithium battery pack, Load power P load Inverter output power P e The angular frequency ω of the VSG is measured.

[0122] Step S4: Determine whether the adjustable power of the current lithium battery pack meets the system power response deficit based on the energy storage response lag criterion; if it meets the energy storage response lag criterion, start timer t2 and proceed to step S5; otherwise, proceed to step S7.

[0123] Step S5: Determine the duration τ of the inertial response phase in the VSG frequency event based on the power disturbance level;

[0124] Step S6: Calculate and output the virtual inertia J and damping coefficient D in real time until t1>τ, then execute step S7; the calculation of virtual inertia J and damping coefficient D is as follows:

[0125] J = J0 + J ΔP +J soc

[0126] D = D0 + D ΔP

[0127] Where: J0 is the reference inertia coefficient; D0 is the reference damping coefficient; J ΔP and D ΔP The adaptive J and D components are characterized by the adjustable power of the lithium battery pack; J soc It considers the J component of battery SOC in system transient events; that is: judging the energy storage lag time 0-τ (stage 1), the improved J and D coefficients are calculated as above;

[0128] Step S7: Obtain the deviation ΔP of the VSG output active power, the difference Δω between the virtual angular frequency ω and the rated angular frequency ω0, and the rate of change of angular frequency dω / dt; calculate and output the virtual inertia J and damping coefficient D using the following expressions (adaptive coefficients of J and D after the energy storage hysteresis criterion is not triggered, and after the energy storage hysteresis criterion is triggered and the optimization stage of τ(S) is completed):

[0129] J = J1 + J SOC

[0130] D = D1

[0131] Where: J1 and D1 are the adaptive J and D coefficients without considering SOC;

[0132] Step S8: Determine if t1>T is true. If true, the round ends, where T is the round duration, T = dt; otherwise, update the real-time SOC of the lithium battery pack and the inverter output power P. e And measure the angular frequency ω of the VSG and return to step S7.

[0133] In this embodiment, the energy storage response hysteresis criterion is:

[0134] ΔP Ba ↑↓≤ρΔP ref

[0135]

[0136] P ref =P load

[0137] In the above formula, the left side of the inequality sign represents the adjustable power of the current lithium battery pack, and the right side is the adjustment coefficient multiplied by the load power change, reflecting whether the current fast power response capability of the hybrid energy storage system meets the VSG requirements; ρ is the adjustment coefficient; ΔP ref The reference power change rate of the VSG; These are the reference power of the VSG in the current round and the reference power of the VSG in the previous round, respectively; the reference power P of the VSG. ref The value is equal to the load power P load .

[0138] In this embodiment, the reference power change rate ΔP of the VSG is used. ref Determine the power disturbance level and assess the duration τ of the inertial response phase in a VSG frequency event:

[0139] If ΔP ref If the value is less than 10%, it is considered a small disturbance, and the duration τ of the inertial response phase in the VSG frequency event is set to τ1.

[0140] If 10% ≤ ΔP ref If the value is ≤30%, it is considered a moderate disturbance, and the duration τ of the inertial response phase in the VSG frequency event is set to τ2.

[0141] If ΔP ref If the value is greater than 30%, it is considered a large disturbance, and the duration τ of the inertial response phase in the VSG frequency event is set to τ3.

[0142] Where τ1 < τ2 < τ3, for example, τ1 = 0.05S, τ2 = 0.1S, τ3 = 0.2S.

[0143] The deviation of the active power output of the VSG is ΔP = (P m -P e Since the angular frequency change rate dω / dt has the same sign in each oscillation stage, ΔP and dω / dt have the same trend. Utilizing this principle, existing technologies use ΔP to replace the derivative term of the angular frequency and correlate it with the J and D coefficients, thereby avoiding the introduction of dω / dt and reducing the impact on frequency and power response characteristics. However, the calculation of ΔP is limited by P. e Real-time calculation, ΔP = P ref -P e P in eThe actual output power needs to be measured in real time by sensors and undergo sampling, filtering, and calculation processes. This process introduces millisecond-level time delays, especially when power fluctuates rapidly, ΔP cannot accurately reflect the instantaneous state changes of the system. The methods mentioned above adjust the J and D coefficients only after the system has shown a frequency deviation trend, and the sampling and calculation delays are superimposed after the event occurs. Considering the adjustment lag that may be caused by the delay, using dω / dt, ΔP, and adaptive J and D coefficients may result in the control being unable to immediately suppress the degree and rate of frequency deviation in stage 1. Therefore, this invention introduces the fuel cell power response coefficient F (average power response per second) and the time constant τ (the duration of stage 1 with moderate disturbance under reference J and D coefficients). By establishing the relationship between the current adjustable power of the lithium battery pack and the active power deviation ΔP, it indirectly correlates with the J and D coefficients, adjusting the appropriate J and D coefficients before the frequency event is detected, avoiding untimely suppression due to factors such as delay.

[0144] In this embodiment, the calculation of the adaptive J and D components, characterized by the adjustable power of the lithium battery pack, is as follows:

[0145]

[0146] Where: F is the power response coefficient of the fuel cell; θ1 represents the adjustment coefficients for the J and D components, respectively. Through the adaptive J and D component calculation method of this invention, suitable J and D coefficients can be obtained at the beginning of the calculation cycle by referencing changes in active power, effectively suppressing... Figure 6 The rate of change and offset of the frequency in stage 1 of (a) are such that the components of the J and D coefficients in this part only act within τ(S) after the hysteresis criterion is triggered.

[0147] Considering the SOC limitation of lithium battery packs, as fast-response energy storage units responsible for responding to transient events in hybrid energy storage systems, when the SOC of the lithium battery pack is at or outside the normal range, transient events may lead to overcharging or over-discharging. Since the power balance equation constraint has higher priority than other constraints such as SOC constraint, the SOC safety of the lithium battery pack is difficult to guarantee when transient events occur. Therefore, this invention considers the relationship between SOC safety and the inertia coefficient J under this condition, and mitigates the deep charge / discharge behavior of the lithium battery pack under this condition by appropriately adjusting the inertia coefficient J.

[0148] In this embodiment, the calculation of the J component of the battery SOC in the system transient event is considered as follows:

[0149]

[0150] in: The adjustment coefficient; SOCref This is a reference SOC for lithium battery packs.

[0151] In this embodiment, the calculation of the adaptive J and D coefficients of SOC is not considered, as follows:

[0152]

[0153] Where: K0 is the threshold value for the change of the difference Δω between the virtual angular frequency and the rated angular frequency during J parameter adjustment; K1 is the threshold value for the change of dω / dt during J parameter adjustment; K D The threshold for the change of Δω during the D parameter adjustment process; k1~k5 are the adjustment coefficients; h is the conversion coefficient between the rate of change of angular frequency and the power deviation; J1 and D1 represent the conventional adaptive J and D adjustment method, without including SOC considerations and energy storage lag considerations.

[0154] In summary, this invention considers the impact of the response delay characteristics of fuel cell hybrid energy storage systems on islanded VSG systems. It uses an energy storage power response lag criterion to determine whether the instantaneous power response capability of the energy storage system meets the demand during sudden changes in load power demand. This invention proposes a pre-adjustment method for inertia and damping coefficients, adaptively adjusting the inertia and damping coefficients in the initial stage of active-frequency events based on the instantaneous power response capability of the energy storage system. The fuel cell hybrid energy storage system proposed in this invention optimizes the cost of energy storage system operation, incorporating not only the main fuel consumption cost but also the degradation degree of the fuel cell and lithium battery, while refining the degradation calculation of the fuel cell.

[0155] The above are preferred embodiments of the present invention. Any changes made to the technical solution of the present invention that do not exceed the scope of the technical solution of the present invention shall fall within the protection scope of the present invention.

Claims

1. A method for controlling ammonia-hydrogen fuel cell hybrid energy storage that optimizes operating costs and islanding stability, characterized in that, Based on an islanded fuel cell-lithium battery hybrid energy storage system, the power allocation with the lowest operating cost that satisfies the constraints of the hybrid energy storage system is solved in real time. An adaptive VSG control strategy that considers both the response delay capability of the energy storage side and the SOC limit of the lithium battery pack is adopted to perform grid-type adaptive VSG control on the inverter in the islanded fuel cell-lithium battery hybrid energy storage system.

2. The method for controlling ammonia-hydrogen fuel cell hybrid energy storage with optimized operating costs and islanding stability according to claim 1, characterized in that, The isolated fuel cell-lithium battery hybrid energy storage system includes an ammonia-hydrogen fuel cell, a power control module, a lithium battery pack, a voltage control module, an inverter, an adaptive VSG control system, an energy management system, and a load. The ammonia-hydrogen fuel cell is connected to the inverter via a Boost converter and employs power control. The lithium battery pack is connected to the inverter via a bidirectional DC-DC module and is controlled by voltage. The energy management system calculates the optimal power allocation based on the input data, system constraints, and system operating costs, and outputs the fuel cell reference power P. fc The power control module; the input data of the energy management system includes the inverter output power P. e Real-time SOC of lithium battery pack, output voltage of lithium battery pack (V) ba Fuel cell output voltage V fc Fuel cell output power P' fc ; Inverter output power P e The calculation is as follows: P e =P m -ω·J dω / dt-ω·D·(ω-ω0) Where: P m Let J be the mechanical power of VSG, and D be the virtual inertia and damping coefficient, respectively. ω and ω0 be the virtual angular frequency and the rated angular frequency, respectively. The power control module is based on the input fuel cell reference power P. fc Output PWM signal for power control of ammonia-hydrogen fuel cell; The voltage control module outputs a PWM signal based on the input DC bus voltage for lithium battery pack voltage control; the DC bus voltage refers to the voltage value of the common DC link connecting the output terminal of the bidirectional DC-DC module and the DC input terminal of the inverter. The inverter is controlled by an adaptive VSG control system, and the output is filtered by an LC filter before supplying power to the load. The adaptive VSG control system includes an adaptive J and D adjustment module, which is used to adjust the coefficients J and D according to the energy storage status, load demand, and AC side stability.

3. The method for controlling ammonia-hydrogen fuel cell hybrid energy storage with optimized operating costs and islanding stability according to claim 1, characterized in that, The quadratic sequential programming algorithm SQP is used to solve for the power allocation with the lowest operating cost that satisfies the constraints of the hybrid energy storage system at each time interval dt. System constraints include power balance equation constraints, fuel cell current change rate constraints, lithium battery pack current change rate constraints, fuel cell power upper limit constraints, and lithium battery pack power upper and lower limit constraints. The system operating cost includes fuel consumption cost, fuel cell degradation cost, lithium battery pack degradation cost, lithium battery pack SOC maintenance cost, lithium battery pack SOC gradient limit penalty cost, and lithium-ion battery pack adjustable power constraint cost. The adjustable power constraint cost of the lithium-ion battery pack is inversely proportional to the adjustable power of the lithium-ion battery pack.

4. The ammonia-hydrogen fuel cell hybrid energy storage control method for optimizing operating costs and islanding stability according to claim 3, characterized in that, The specific calculations for fuel consumption costs, fuel cell degradation costs, lithium battery pack degradation costs, and lithium battery pack SOC maintenance costs are as follows: Fuel consumption cost Where: λ H2 This represents the fuel consumption cost coefficient; m H2 M represents the hydrogen fuel consumption cost of the system during the current time interval dt; H2 The market price of hydrogen fuel; N is the number of individual cells in the fuel cell; F is the Faraday constant; I fc (t) represents the fuel cell output current, I fc =P′ fc / V fc ; Fuel cell degradation costs Where: λ fc For fuel cell degradation cost coefficient; SOD b (t) is the degradation formula for the fuel cell, representing the degree of degradation of the fuel cell at the current fuel cell power level; c0, c1, and c2 are the coefficients of the quadratic function; P′ fc (t) represents the fuel cell output power at time t; x fc This refers to the output voltage of the fuel cell; Degradation cost of lithium battery packs Where: λ ba ΔQ is the degradation cost coefficient for lithium battery packs. loss The degree of degradation of the lithium battery pack within the current time interval dt; a, b, c, d′, e′ are empirical coefficients; T e Absolute temperature; Ah t0 For throughput per hour; I rate For charge / discharge rate, I rate =P ba / (V ba ·Q ba ), Q ba P is the rated capacity of the lithium battery pack. ba =P e -P′ fc P e This refers to the inverter's output power. Lithium-ion battery pack SOC maintenance cost: Where: λ SOC To maintain the cost factor of lithium battery pack SOC, These are the hydrogen consumption cost, fuel cell degradation cost, and lithium battery pack degradation cost, respectively, incurred by the system in maintaining the state of charge (SOC) of the lithium battery pack.

5. The method for controlling ammonia-hydrogen fuel cell hybrid energy storage with optimized operating costs and islanding stability according to claim 1, characterized in that, Adaptive VSG control specifically involves executing the following steps in cycles with a time interval dt: Step S1: Start the round and start timer t1; Step S2: Parameter initialization; Step S3: Obtain the adjustable power ΔP of the current lithium battery pack. Ba ↑↓, Real-time SOC of lithium battery pack, Load power P load Inverter output power P e The angular frequency ω of the VSG is measured. Step S4: Determine whether the adjustable power of the current lithium battery pack meets the system power response deficit based on the energy storage response lag criterion; if it meets the energy storage response lag criterion, start timer t2 and proceed to step S5; otherwise, proceed to step S7. Step S5: Determine the duration τ of the inertial response phase in the VSG frequency event based on the power disturbance level; Step S6: Calculate and output the virtual inertia J and damping coefficient D in real time until t1>τ, then execute step S7; the calculation of virtual inertia J and damping coefficient D is as follows: J=J0+J ΔP +J soc D=D0+D ΔP Where: J0 is the reference inertia coefficient; D0 is the reference damping coefficient; J ΔP and D ΔP The adaptive J and D components are characterized by the adjustable power of the lithium battery pack; J soc This considers the J component of battery SOC in system transient events; Step S7: Obtain the deviation ΔP of the VSG output active power, the difference Δω between the virtual angular frequency ω and the rated angular frequency ω0, and the rate of change of angular frequency dω / dt; calculate and output the virtual inertia J and damping coefficient D using the following expressions: J=J1+J SOC D = D1 Where: J1 and D1 are the adaptive J and D coefficients without considering SOC; Step S8: Determine if t1>T is true. If true, the round ends, where T is the round duration, T = dt; otherwise, update the real-time SOC of the lithium battery pack and the inverter output power P. e And measure the angular frequency ω of the VSG and return to step S7.

6. The method for controlling ammonia-hydrogen fuel cell hybrid energy storage with optimized operating costs and islanding stability according to claim 1, characterized in that, The energy storage response hysteresis criterion is as follows: ΔP Ba ↑↓≤ρΔP ref P ref =P load Where: ρ is the adjustment coefficient; ΔP ref The reference power change rate of the VSG; These are the reference power of the VSG in the current round and the reference power of the VSG in the previous round, respectively; the reference power P of the VSG. ref The value is equal to the load power P load .

7. The method for controlling ammonia-hydrogen fuel cell hybrid energy storage with optimized operating costs and islanding stability according to claim 5, characterized in that, Using the reference power change rate ΔP of VSG ref Determine the power disturbance level and assess the duration τ of the inertial response phase in a VSG frequency event: If ΔP ref If the value is less than 10%, it is considered a small disturbance, and the duration τ of the inertial response phase in the VSG frequency event is set to τ1. If 10% ≤ ΔP ref If the value is ≤30%, it is considered a moderate disturbance, and the duration τ of the inertial response phase in the VSG frequency event is set to τ2. If ΔP ref If the value is greater than 30%, it is considered a large disturbance, and the duration τ of the inertial response phase in the VSG frequency event is set to τ3. Among them, τ1<τ2<τ3.

8. The method for controlling ammonia-hydrogen fuel cell hybrid energy storage with optimized operating costs and islanding stability according to claim 5, characterized in that, The calculation of the adaptive J and D components, characterized by the adjustable power of the lithium battery pack, is as follows: Where: F is the power response coefficient of the fuel cell; θ1 are the adjustment coefficients for the J and D components, respectively.

9. The method for controlling ammonia-hydrogen fuel cell hybrid energy storage with optimized operating costs and islanding stability according to claim 5, characterized in that, The calculation of the J component of battery SOC during system transient events is as follows: in: The adjustment coefficient; SOC ref This is a reference SOC for lithium battery packs.

10. The ammonia-hydrogen fuel cell hybrid energy storage control method for optimizing operating costs and islanding stability according to claim 5, without considering the calculation of adaptive J and D coefficients of SOC, is as follows: in: K0 is the threshold value for the change of the difference Δω between the virtual angular frequency and the rated angular frequency during J parameter adjustment; K1 is the threshold value for the change of dω / dt during J parameter adjustment; K D Δω represents the threshold value for the change in D parameter during adjustment; k1 to k5 are adjustment coefficients; h is the conversion coefficient between the rate of change of angular frequency and the power deviation.

Images