Method and system for predicting load current of fuel cell and vehicle

Abstract

The invention belongs to the technical field of battery management, and particularly relates to a fuel cell load current prediction method and system and a vehicle, and the method comprises the steps: constructing a state space model of a fuel cell hybrid power system, the charge state of the lithium battery and the output power of the fuel cell are used as state quantities; and setting an objective function and a constraint condition of energy management for the constructed state space model, and solving a minimization problem based on the objective function and the constraint condition to obtain a control input quantity enabling the objective function to be minimum, namely determining the optimal fuel cell load current. The invention can improve the operation efficiency of the fuel cell and ensure the efficient and stable operation, and can achieve the effects of stabilizing the output of the cell and reducing the loss of the cell.

Description

Technical Field

[0001] The present invention belongs to the technical field of energy management, and specifically relates to a method, a system and a vehicle for predicting the load current of a fuel cell. Background Art

[0002] With the increasingly severe global energy crisis and environmental pollution problems, new energy vehicles, as an important alternative to traditional fuel vehicles, are gradually becoming the mainstream trend in the development of the automotive industry. Among the many new energy vehicle technology routes, fuel cell-powered vehicles are regarded as one of the new energy vehicle types with the greatest development potential due to their high efficiency, zero emissions or low emissions. As the core power source of fuel cell-powered vehicles, fuel cells can not only achieve rapid energy replenishment (similar to the refueling experience of traditional fuel vehicles), but also inherit the advantages of clean energy of new energy vehicles, which is of great significance for promoting green travel and reducing carbon emissions.

[0003] However, the development of fuel cell vehicles still faces many technical challenges, one of which is how to efficiently and stably manage the energy output of fuel cells. There is a complex relationship between the efficiency of fuel cells and their output power, which directly affects the dynamic performance, driving range of the whole vehicle and the service life of the battery.

[0004] Therefore, it is necessary to develop a new method, a system and a vehicle for predicting the load current of a fuel cell. Summary of the Invention

[0005] The purpose of the present invention is to provide a method, a system and a vehicle for predicting the load current of a fuel cell, which can improve the operating efficiency of the fuel cell and ensure its efficient and stable operation, so as to achieve a stable battery output and reduce battery loss.

[0006] In a first aspect, an energy management method of a battery management system according to the present invention includes the following steps:

[0007] Construct a state space model of a fuel cell hybrid system, where the state space model takes the fuel cell load current as a control input variable and the state of charge of the lithium battery and the output power of the fuel cell as state variables;

[0008] Set an objective function and constraint conditions for energy management for the constructed state space model, and by solving the minimization problem based on the objective function and constraint conditions, obtain the control input variable that minimizes the objective function, that is, determine the optimal fuel cell load current.

[0009] Optionally, configuring the objective function and constraint conditions for the constructed state space model specifically includes:

[0010] Construct an objective function for energy management, which is used to measure the performance of the system over an infinite time horizon starting from the current time;

[0011] Define an energy function, which is used to measure the energy or stability of the system state from the current time to a future time;

[0012] Based on the objective function and the energy function, combined with the system dynamic characteristics and stability requirements, construct the constraint conditions for solving the minimization problem. The objective function covers the infinite time horizon performance evaluation, enabling the algorithm to take into account the long-term behavior and overall performance of the system, optimize the energy utilization efficiency, and avoid local optima; secondly, the energy function accurately quantifies the energy state and stability of the system, providing a more practical optimization basis for the algorithm; finally, the constraint conditions constructed by combining the objective function, the energy function, and the system dynamics and stability requirements ensure that the solution process not only conforms to the actual operation law of the system but also guarantees the stable and safe operation of the system.

[0013] Optionally, based on the objective function and the energy function, combined with the system dynamic characteristics and stability requirements, construct the constraint conditions for solving the minimization problem, specifically:

[0014] Construct an inequality based on the objective function and the energy function;

[0015] Sum the inequality from 0 to ∞ to obtain the inequality relationship between the energy function at the current time and the infinite time horizon cost function;

[0016] Define an upper bound of the performance function. Based on this upper bound of the performance function and the inequality relationship, combined with the Lyapunov stability theorem, obtain the constraint conditions for solving the minimization problem. By constructing and summing the inequality based on the objective function and the energy function, the comprehensive consideration of energy and cost from the current to the infinite time horizon is realized, enhancing the rigor of the mathematical model and making the optimization more practical; by summing to establish a global association, the algorithm can globally optimize the system performance and avoid local optima; defining the upper bound of the performance function and combining with the Lyapunov stability theorem ensure the dynamic stability of the system during the optimization process and prevent the state from diverging.

[0017] Optionally, the state space model of the fuel cell hybrid system is:

[0018] x(k + 1) = A(k)x(k) + B(k)u(k)

[0019] y(k) = C(k)x(k)

[0020] Among them, \(x(k + 1)\) represents the state quantity at time \(k + 1\); \(x(k)\) represents the state quantity at time \(k\); \(A(k)\) is the state matrix related to time \(k\); \(B(k)\) is the control input matrix related to time \(k\); \(u(k)\) represents the control input quantity, and \(u(k)=I st (k)\), where \(I st (k)\) is the fuel cell load current; \(y(k)\) represents the output quantity; \(C(k)\) is the output matrix. This provides a specific mathematical model for subsequent algorithm design and simulation verification.

[0021] Optionally, the infinite-horizon cost function is:

[0022]

[0023] Among them, \(J ∞ (k)\) represents the infinite-horizon cost function, which is used to measure the performance of the system starting from the current time \(k\) within an infinite future time horizon. The smaller the value of \(J ∞ (k)\), the closer the performance of the system is to the expectation; \(x(k + i|k)\) represents the prediction of the system state at future time \(k + i\) at time \(k\); \(u(k + i|k)\) represents the prediction of the control input to be applied at future time \(k + i\) at time \(k\); \(L\) is a constant matrix of system state weights; \(R\) is a constant matrix of control input quantity weights; \(T\) represents the transpose. By adjusting the constant matrix \(L\) of system state weights and the constant matrix \(R\) of control input quantity weights, the importance that the system attaches to future performance and current control input can be flexibly adjusted, thereby achieving different control objectives.

[0024] Optionally, the energy function is:

[0025] \(V(x(k + i|k)) = x(k + i|k) T \cdot P\cdot x(k + i|k)

[0026] Among them, \(V(x(k + i|k))\) represents the energy function, which is used to measure the energy or stability of the state \(x(k + i|k)\) of the system at future time \(k + i\) at time \(k\), and \(P\) is the Lyapunov matrix. By introducing the Lyapunov matrix \(P\), the positive definiteness of the energy function can be ensured, thus providing a powerful tool for the stability analysis of the system.

[0027] Optionally, the inequality relationship between the energy function at the current time and the infinite-horizon cost function is:

[0028] \(-V(x(k|k))\leq -J ∞ (k)\).

[0029] Among them, V(x(k|k)) is the energy function at the current time. The inequality relationship between the energy function at the current time and the infinite-time domain cost function is given, providing a key constraint for subsequent solution.

[0030] Optionally, the constraint condition is:

[0031]

[0032] Among them, A = A(k); B = B(k); I is the standard matrix; Q is the performance constraint matrix; Y = FQ, and F is the control gain matrix. It ensures the system stability, optimizes the control performance, and ensures the efficient operation of the energy management strategy under complex working conditions.

[0033] In the second aspect, a prediction system for the load current of a fuel cell according to the present invention includes a controller and a memory. A computer-readable program is stored in the memory. When the computer-readable program is called by the controller, it can execute the steps of the prediction method for the load current of the fuel cell according to the present invention.

[0034] In the third aspect, a vehicle according to the present invention adopts an energy management system of the battery management system according to the present invention.

[0035] Advantages of the present invention: Aiming at the problem of the relationship between the battery efficiency and the battery output power in a fuel cell, the present invention adopts a predictive control algorithm to deeply optimize the energy management strategy of a fuel cell hybrid vehicle. This method performs control solution under the condition of meeting actual physical constraints, realizes the optimal control effect, improves the operating efficiency of the fuel cell and ensures its efficient and stable operation, achieves a stable battery output while reducing battery loss and enhancing the discharge efficiency of the fuel cell. Description of the Drawings

[0036] Figure 1 It is a flowchart of the prediction method for the load current of the fuel cell in the embodiment of the present application;

[0037] Figure 2 It is a schematic structural diagram of a fuel cell hybrid system in the embodiment of the present application;

[0038] Figure 3 It is a schematic diagram of the internal resistance model of a lithium battery in the embodiment of the present application;

[0039] Figure 4 It is a strategy diagram of the energy management of the battery management system in the embodiment of the present application;

[0040] Figure 5 It is a principle block diagram of the prediction system for the load current of the fuel cell in the embodiment of the present application. Detailed Embodiments

[0041] The embodiments of the present invention will be described below with reference to the accompanying drawings and preferred embodiments. Those skilled in the art can understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through different specific embodiments. Various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be understood that the preferred embodiments are only for illustrating the present invention, rather than for limiting the protection scope of the present invention.

[0042] As Figure 1 shown, in the embodiment of the present application, a method for predicting the load current of a fuel cell includes the following steps:

[0043] Construct a state - space model of the fuel - cell hybrid power system, wherein the state - space model takes the fuel - cell load current as the control input quantity, and the state of charge of the lithium battery and the output power of the fuel cell as the state variables.

[0044] Set the objective function and constraint conditions for energy management for the constructed state - space model, and by solving the minimization problem based on the objective function and constraint conditions, obtain the control input quantity that minimizes the objective function, that is, determine the optimal fuel - cell load current.

[0045] This method aims at the complex relationship between the battery efficiency and the output power in the fuel cell, and adopts a predictive control algorithm to optimize the energy management strategy of the fuel - cell hybrid vehicle. This predictive control algorithm performs fine - grained solution under actual physical constraint conditions, aiming to achieve the optimal control effect. Through this strategy, not only can the operating efficiency of the fuel cell be significantly improved, ensuring its efficient and stable operation, but also the smoothness of the battery output can be achieved, effectively reducing battery loss, and further improving the discharge efficiency of the fuel cell, thus comprehensively optimizing the overall performance of the vehicle.

[0046] In a possible embodiment, configuring the objective function and constraint conditions for the constructed state - space model specifically includes:

[0047] Construct an objective function for energy management, and the objective function is used to measure the performance of the system in the infinite future time domain starting from the current time.

[0048] Define an energy function, and the energy function is used to measure the energy or stability of the state of the system from the current time to the future time.

[0049] Based on the objective function and the energy function, combined with the system dynamic characteristics and stability requirements, construct the constraint conditions for solving the minimization problem.

[0050] In a possible embodiment, based on the objective function and the energy function, combining the system dynamic characteristics and stability requirements, a constraint condition for solving the minimization problem is constructed, specifically:

[0051] Construct an inequality based on the objective function and the energy function.

[0052] Sum the inequality from 0 to ∞ to obtain the inequality relationship between the energy function at the current time and the infinite-horizon cost function.

[0053] Define an upper bound of the performance function. Based on this upper bound of the performance function and the inequality relationship, combined with the Lyapunov stability theorem, a constraint condition for solving the minimization problem is obtained.

[0054] The steps of the method for predicting the load current of the fuel cell in the embodiments of the present application are described in detail below:

[0055] 1. Construct a state-space model of the fuel cell hybrid system:

[0056] Based on the principle analysis of the fuel cell hybrid system and the internal resistance model of the lithium battery, a state-space model of the fuel cell hybrid system is constructed, and the state variables, control input variables, and outputs of the fuel cell hybrid system are defined. Among them, the state variables include the state of charge of the lithium battery and the output power of the fuel cell, the control input variable is the load current of the fuel cell, and the load current of the fuel cell is determined by the hydrogen consumption rate in the fuel cell and the system efficiency. Among them, k is the sampling time. The outputs are the state of charge of the lithium battery and the output power of the fuel cell.

[0057] Among them, for the structure of the fuel cell hybrid system, see Figure 2 , the output end of the fuel cell is connected to a unidirectional DC-DC converter, the unidirectional DC-DC converter is connected to the bus, and the lithium battery is directly connected to the bus. Under this structure, the terminal voltage of the lithium battery changes with the bus voltage, the terminal voltage and load current of the fuel cell are controlled by the unidirectional DC-DC converter, and at the same time, due to the existence of the unidirectional DC-DC converter, the harm of bus current reverse injection is also avoided. Since there are no intermediate components, the lithium battery can provide a faster power response and higher energy utilization rate.

[0058] For the internal resistance model of the lithium battery, see Figure 3 , the lithium battery is equivalent to a series circuit of a voltage source V oc and a resistor R int , and the terminal voltage and state of charge (SOC) of the lithium battery during charging and discharging are calculated.

[0059] The state-space model of the fuel cell hybrid system is:

[0060] x(k + 1) = A(k)x(k) + B(k)u(k)

[0061] y(k) = C(k)x(k)

[0062] Among them, x(k + 1) represents the state quantity at time k + 1; x(k) represents the state quantity at time k; A(k) is the state matrix related to time k; B(k) is the control input matrix related to time k; u(k) represents the control input quantity, and u(k) = I st (k), and I st (k) is the fuel cell load current; y(k) represents the output quantity; C(k) is the output matrix. It provides a specific mathematical model for subsequent algorithm design and simulation verification.

[0063] The construction process of the state - space model of the fuel - cell hybrid power system is as follows:

[0064] Establish the state - space equation:

[0065]

[0066] Among them: SOC(k + 1) is the state of charge of the lithium - battery at time k + 1; SOC(k) is the state of charge of the lithium - battery at time k; P v (k) is the required power of the whole vehicle at time k; P fc (k) is the output power of the fuel cell at time k; T is the sampling period; V BP (k) is the terminal voltage of the lithium - battery at time k; Q C is the capacity of the lithium - battery; P fc (k + 1) is the output power of the fuel cell at time k + 1; H(I st (k)) is a function of I st (k), representing the derivative of power with respect to current multiplied by the current change.

[0067]

[0068] P v (k) - P fc (k): represents the charging and discharging power of the lithium - battery. When P v (k) - P fc (k) is positive, it represents discharging. When P v (k) - P fc (k) is negative, it represents charging.

[0069] (P v (k) - P fc (k))T: represents the energy change.

[0070] The matrix form of the state - space equation:

[0071] x(k + 1) = A(k)x(k) + B(k)u(k)

[0072] where denotes defined equal to, T denotes transpose. A(k) and B(k) can be obtained by substituting the parameters in the system

[0073] 2. Set the objective function and constraints of energy management for the constructed state - space model:

[0074] When designing the energy management strategy, it is necessary to consider not only from the economic perspective but also from the perspectives of maintaining the SOC of the power battery and the efficiency of the fuel cell

[0075] First, establish the infinite - horizon cost function (i.e., the objective function of energy management):

[0076]

[0077] where J ∞ (k) is the infinite - horizon cost function, used to measure the performance of the system starting from the current time k in the future infinite - horizon. The smaller the value of J ∞ (k), the closer the performance of the system is to the expectation. x(k + i|k) is the state prediction value, representing the prediction of the system state at future time k + i at time k. u(k + i|k) is the control - input prediction value, representing the prediction of the control input to be applied at future time k + i at time k. L represents the constant matrix of the system - state weight, and R represents the constant matrix of the control - input weight. Among them, L and R are constant matrices determined based on the actual situation

[0078] Define the energy function as:

[0079] V(x(k + i|k)) = x(k + i|k) T ·P·x(k + i|k)

[0080] where V(x(k + i|k)) represents the energy function, used to measure the energy or stability of the system state x(k + i|k) at future time k + i at time k; P is the Lyapunov matrix

[0081] Next, construct inequalities based on the objective function of energy management and the energy function. The specific inequalities are as follows:

[0082] V(x(k + i + 1|k)) - V(x(k + i|k)) ≤ -((x(k + i|k) T )·L·x(k + i|k) + (u(k + i|k) T ·R·

[0083] u(k+i|k)))

[0084] Sum the inequality from 0 to ∞:

[0085]

[0086] Furthermore, an inequality relationship between the energy function at the current time and the infinite-horizon cost function is obtained:

[0087] -V(x(k|k)) ≤ -J ∞ (k)

[0088] where V(x(k|k)) represents the energy function at the current time.

[0089] Let the energy function have an upper bound, i.e., V(x(k|k)) ≤ γ, where γ is the upper bound of the performance function;

[0090] Since V(x(k|k)) ≤ γ, so we have:

[0091] J ∞ (k) ≤ γ

[0092] Substitute the energy function into the upper bound of the performance function, and we get:

[0093] x(k+i|k) T ·P·x(k+i|k) ≤ γ

[0094] Construct a quadratic inequality:

[0095] To transform the above inequality into matrix form, a quadratic inequality can be considered.

[0096] First, rewrite the quadratic inequality as:

[0097] γ - x(k+i|k) T ·P·x(k+i|k) ≥ 0

[0098] Rewrite it further as:

[0099]

[0100] where Q is the performance constraint matrix,

[0101] Based on Lyapunov stability theorem, the constraint conditions for solving the minimization problem are obtained:

[0102]

[0103] where A is A(k), B is B(k), I is the identity matrix; Y = FQ, and F is the control gain matrix.

[0104] Finally, solve the minimization problem:

[0105] According to the decision variables Q, Y, and γ, it is transformed into solving Obtain the energy management algorithm of the battery management system, which is specifically as follows:

[0106]

[0107] Finally, obtain the optimal system input u(k + 1) that minimizes the infinite-horizon cost function, u(k + 1) = Fx(k + 1).

[0108] Through this method, the future dynamic behavior of the system under a certain control action can be predicted. On this basis, the optimal control input quantity is solved recursively according to the given constraint conditions and performance requirements, and the fuel cell hybrid system is controlled based on this control input quantity. At each step of the recursion, the prediction of the future dynamic behavior is corrected by obtaining the actual state quantity.

[0109] This method can be summarized into three principles: model construction, rolling optimization, and feedback correction.

[0110] Model construction: Use the state equation as the prediction model.

[0111] Rolling optimization: At each sampling time, the optimization index only covers the future finite horizon starting from this sampling time. Therefore, it is an open-loop optimization problem with the future finite control quantities as the optimization variables. After obtaining the optimal control input quantity, predictive control does not implement them one by one, but only applies the current control input quantity to the fuel cell hybrid system. By the next sampling time, this optimization horizon rolls forward simultaneously as time advances. Therefore, predictive control does not adopt a global optimization performance index, but has an optimization performance index relative to that time at each time. The relative forms of the optimization performance indices at different times are the same, but the specific time intervals they contain are different. This shows that the optimization in predictive control is not carried out offline once, but repeatedly online.

[0112] Feedback correction: The rolling optimization based on the prediction model is only open-loop optimization. Due to various uncertainties such as prediction model mismatch and unknown disturbances that inevitably exist in the actual system, the actual operation of the system may deviate from the ideal optimization result. To compensate for the influence of various uncertainty factors on the system to a certain extent, a closed-loop mechanism is introduced. At each sampling time, first, the real-time state or output information of the object is detected, and before solving the control action by optimization, this feedback information is used to refresh or correct the next-step prediction and optimization based on the actual situation. For the prediction control algorithm based on the state equation, the measured system state can be directly used as the new basis point for each-step prediction and optimization without additional correction. It can be seen that although the prediction control performs open-loop optimization at each step, due to the combination of feedback correction, the entire rolling process realizes closed-loop optimization.

[0113] Among them, when the vehicle is powered on, the state of charge of the lithium battery is collected, and the control target (i.e., the output power of the fuel cell) is determined. After the vehicle is powered on, the specific values of A(k), B(k), and C(k) in the state-space model are calculated in real time according to the data collected by the vehicle, and are substituted into the objective function and energy function of the state-space model. By solving the minimization problem, the system input for the next sampling period, that is, the load current of the fuel cell, is obtained in real time and input to the fuel cell hybrid system for execution.

[0114] Such as Figure 4 shown, on the basis of meeting the vehicle demand power, the energy distribution relationship between the lithium battery and the fuel cell is obtained in real time by the controller. When it is executed for the first time, before the vehicle is powered on, the basic parameters of the controller also need to be determined according to the operating conditions of the system in actual applications, including the sampling period, the weight of the system objective function, the target holding value of the state of charge, and the output power with the highest fuel cell efficiency. When it is not the first execution, this step (i.e., determining the basic parameters of the controller) is not executed, and it directly starts to execute from the step of determining the control target.

[0115] Such as Figure 5 shown, in the embodiment of the present application, a prediction system for the load current of a fuel cell includes a controller and a memory. When the computer-readable program stored in the memory is called by the controller, it can execute the steps of the method for predicting the load current of the fuel cell in the embodiment of the present application.

[0116] In the embodiment of the present application, a vehicle adopts an energy management system such as the battery management system in the embodiment of the present application.

[0117] The above embodiments are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above embodiments. Any other changes, modifications, substitutions, combinations, and simplifications made without departing from the spirit and principle of the present invention shall be equivalent replacement methods and are all included in the protection scope of the present invention.

Claims

1. A method for predicting the load current of a fuel cell, characterized in that Including the following steps: Construct a state - space model of the fuel - cell hybrid power system, where the state - space model uses the fuel - cell load current as the control input quantity, and the state of charge of the lithium battery and the output power of the fuel cell as the state variables; Set the objective function and constraint conditions for energy management for the constructed state - space model. By solving the minimization problem based on the objective function and constraint conditions, obtain the control input quantity that minimizes the objective function, that is, determine the optimal fuel - cell load current.

2. The method for predicting the load current of a fuel cell according to claim 1, wherein Configure the objective function and constraint conditions for the constructed state - space model, specifically: Construct an objective function for energy management, which is used to measure the performance of the system in the infinite future time domain starting from the current time; Define an energy function, which is used to measure the energy or stability of the state of the system from the current time to the future time; Based on the objective function and the energy function, combined with the system dynamic characteristics and stability requirements, construct constraint conditions for solving the minimization problem.

3. The method for predicting the load current of a fuel cell according to claim 2, characterized in that, Based on the objective function and the energy function, combined with the system dynamic characteristics and stability requirements, construct constraint conditions for solving the minimization problem, specifically: Construct an inequality based on the objective function and the energy function; Sum the inequality from 0 to ∞ to obtain the inequality relationship between the energy function at the current time and the infinite - time - domain cost function; Define an upper bound of the performance function. Based on this upper bound of the performance function and the inequality relationship, combined with the Lyapunov stability theorem, obtain the constraint conditions for solving the minimization problem.

4. The method for predicting the load current of a fuel cell according to claim 3, wherein The state - space model of the fuel - cell hybrid power system is: x(k + 1)=A(k)x(k)+B(k)u(k) y(k)=C(k)x(k) Among them, \(x(k + 1)\) represents the state quantity at time \(k + 1\); \(x(k)\) represents the state quantity at time \(k\); \(A(k)\) is the state matrix related to time \(k\); \(B(k)\) is the control input matrix related to time \(k\); \(u(k)\) represents the control input quantity, \(u(k)=I st (k)\), \(I st (k)\) is the fuel cell load current; \(y(k)\) represents the output quantity; \(C(k)\) is the output matrix.

5. The method for predicting the load current of a fuel cell according to claim 4, wherein The objective function adopts an infinite - time - domain cost function, specifically: Among them, J ∞ (k) represents the infinite-horizon cost function, which is used to measure the performance of the system starting from the current time k over an infinite future time horizon. The smaller the value of J ∞ (k), the closer the performance of the system is to the expectation; x(k+i|k) represents the prediction of the system state at future time k+i at time k; u(k+i|k) represents the prediction of the control input to be applied at future time k+i at time k; L represents the constant matrix of the system state weight; R represents the constant matrix of the control input quantity weight; and T represents the transpose.

6. The method for predicting the load current of a fuel cell according to claim 5, wherein The energy function is: V(x(k+i|k)) = x(k+i|k) T ·P·x(k+i|k) Among them, V(x(k + i|k)) represents the energy function, which is used to measure the energy or stability of the state x(k + i|k) of the system at time k for the future time k + i; P is the Lyapunov matrix.

7. The method for predicting the load current of a fuel cell according to claim 6, wherein The inequality relationship between the energy function at the current time and the infinite - time - domain cost function is: -V(x(k|k)) ≤ -J ∞ (k); Among them, V(x(k|k)) is the energy function at the current time.

8. The method for predicting the load current of a fuel cell according to claim 7, wherein The constraint conditions are: Among them, A = A(k); B = B(k); I is the identity matrix; Q is the performance constraint matrix; Y = FQ, and F is the control gain matrix.

9. A prediction system for fuel - cell load current, including a controller and a memory. The memory stores a computer - readable program. When the computer - readable program is called by the controller, it can execute the steps of the fuel - cell load - current prediction method according to any one of claims 1 to 8.

10. A vehicle, characterized in that: Adopt the fuel - cell load - current prediction system according to claim 9.

Images