Energy management method of fuel cell vehicle

Abstract

The invention relates to an energy management method of a fuel cell vehicle, which comprises the following steps: establishing calculation models including a vehicle dynamics model, a fuel cell model and a lithium battery model; dividing the current driving condition of the vehicle into different stages according to the power demand of the vehicle and the gradient change of the power demand, including a high-power rapid power response stage, a high-power slow-speed power response stage, a low-power rapid power response stage and a low-power slow-speed power response stage; for a high-power slow-speed power response stage and a low-power slow-speed power response stage, distributing the power of the fuel cell and the lithium battery by adopting an MPC method based on the calculation model; and for the high-power quick power response stage and the low-power quick power response stage, directly distributing the power of the fuel cell and the lithium battery by adopting a state machine method. According to the invention, global optimization of power distribution of the fuel cell vehicle under various driving conditions is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of new energy power systems, and in particular to an energy management method for a fuel cell vehicle. Background Art

[0002] A proton exchange membrane fuel cell (PEMFC) is a clean energy conversion device that generates electrical and thermal energy through the electrochemical reaction of hydrogen and oxygen. It has now been widely used in devices such as mobile power supplies, automobiles, and drones. Its power generation process produces almost no noise and achieves zero carbon emissions. However, the output characteristics of fuel cells are relatively weak and cannot meet instantaneous high-power demands. Therefore, they usually need to be equipped with auxiliary power sources, such as lithium batteries. A system composed of a fuel cell and a lithium battery forms a PEMFC hybrid power system. Compared with traditional vehicles, a fuel cell vehicle (FCV) equipped with a fuel cell hybrid power system has a better cruising range. However, the highly nonlinear characteristics of the power system composed of multiple energy sources pose design challenges to the energy management strategy (EMS). Especially in terms of adaptability and control robustness under various driving conditions. Under this premise, advanced mechatronics technology plays a considerable role in releasing the energy-saving potential of the hybrid power system.

[0003] Currently, there are two mainstream types of EMS for various hybrid power systems: rule-based EMS and optimization-based EMS. The rule-based EMS uses deterministic or fuzzy logic strategies and has the advantages of simple structure and easy implementation. However, this method heavily relies on expert experience and actual operation constraints and lacks flexibility, especially when facing complex driving conditions. The rule-based method usually distributes power between the fuel cell and the lithium-ion battery according to different power demand stages. This method mainly considers the operation of a single battery and will cause fuel waste to a certain extent. The optimization-based EMS effectively solves the above problems by setting different objectives to constrain and distribute the power between the fuel cell and the lithium-ion battery. Indicators such as the state of charge (SOC) of the lithium battery, the temperature of the battery stack, and the hydrogen mass flow rate are often used as optimization objectives to balance the power distribution in the hybrid power system. The optimization-based EMS aims to obtain the best control sequence based on known driving information. However, this strategy usually does not consider the information of the driving conditions, resulting in the energy management algorithm often only achieving local optimality rather than global optimality, making it difficult to be applied in actual situations. Summary of the Invention

[0004] In view of the deficiencies of the prior art, the present invention provides an energy management method for a fuel cell vehicle, aiming to achieve global optimization of power distribution under various driving conditions of the fuel cell vehicle.

[0005] The technical solution adopted by the present invention is as follows:

[0006] An energy management method for a fuel cell vehicle, wherein the fuel cell vehicle adopts a hybrid power system composed of a fuel cell and a lithium battery, and is characterized in that the energy management method includes:

[0007] Establish a calculation model, which includes a vehicle dynamics model, a fuel cell model, and a lithium battery model;

[0008] According to the power demand of the vehicle and the change of the power demand gradient, the current driving condition of the vehicle is divided into different stages, including a high-power fast power response stage, a high-power slow power response stage, a low-power fast power response stage, and a low-power slow power response stage;

[0009] For the high-power slow power response stage and the low-power slow power response stage, the MPC method is used to distribute the power of the fuel cell and the lithium battery, including:

[0010] Taking the degree of the state of charge of the lithium battery deviating from the reference value being low and minimizing the hydrogen consumption of the fuel cell as the goal, an optimization target model is established;

[0011] By the MPC method, the optimization target model and the calculation model are constructed into a standard quadratic objective function;

[0012] Based on the vehicle dynamics model, the required power at the current moment is calculated;

[0013] Based on the fuel cell model, the fuel cell output power at the current moment is calculated;

[0014] Based on the lithium battery model, the state of charge of the lithium battery at the current moment is calculated;

[0015] Substitute the required power, fuel cell output power, and state of charge of the lithium battery at the current moment into the standard quadratic objective function for solution to obtain the predicted fuel cell output power and lithium battery output power at the next moment, and then according to the calculation model, obtain the fuel cell hydrogen consumption and the state of charge of the lithium battery at the next moment, so as to achieve power distribution;

[0016] For the high-power fast power response stage and the low-power fast power response stage, the state machine method is used to directly distribute the power of the fuel cell and the lithium battery.

[0017] The further technical solution is:

[0018] The optimized objective model is as follows:

[0019] minJ = J1 + J2

[0020] s.t. SOC min ≤ SOC(k + i|k) ≤ SOC max

[0021] 0 ≤ P FC (k + i|k) ≤ P FC,max

[0022] Wherein,

[0023] In the formula, SOC, SOC min , SOC max are respectively the state of charge of the lithium battery, its minimum value and maximum value, k represents the current moment, P FC , P FC,max is the output power of the fuel cell and its maximum value; N p is the prediction interval step size, i represents the i-th time step within N p time steps, p and q are weight coefficients, SOC ref is the reference value of the state of charge of the lithium battery.

[0024] The vehicle dynamics model is as follows:

[0025]

[0026] Wherein, P load , P acc，r , P roll，r , P air，r are respectively the required power of the vehicle, the resistance loss during acceleration, the rolling resistance loss and the air resistance loss, v ve is the driving speed of the vehicle, α roll , α f , α air are respectively the acceleration coefficient, the friction coefficient and the air resistance coefficient, m and a are respectively the vehicle mass and the vehicle acceleration, g is the acceleration due to gravity, A is the frontal area, η t,e is the transmission efficiency of the vehicle motor.

[0027] The fuel cell model includes:

[0028] P FC = N cell V FC I FC

[0029] = N cell (V Nernst-V act -V ohmic -V conc )I FC

[0030] Among them, N cell 、V FC 、I FC are respectively the number of series-connected fuel cells, the actual output circuit voltage, and the external circuit current; V Nernst 、V act 、V ohmic 、V conc are respectively the Nernst open-circuit voltage, activation polarization voltage, ohmic polarization voltage, and concentration polarization voltage;

[0031] Theoretical hydrogen consumption rate calculation formula:

[0032]

[0033] Among them, M H2 is the molar mass of hydrogen, and F is the Faraday constant.

[0034] The lithium battery model includes:

[0035] Lithium battery output power:

[0036]

[0037] Among them, the circuit of the lithium battery is simplified to an ideal resistor and an ideal open-circuit voltage source connected in series with the ideal resistor, P bat 、V ba 、I ba are respectively the lithium battery output power, the voltage of the lithium battery, and the external circuit current, R ba 、V OC are respectively the resistance value of the ideal resistor and the voltage value of the ideal open-circuit voltage source;

[0038] Among them, V OC =a1SOC + a2SOC 2 , where a1 and a2 are constants obtained by fitting according to the actual operation data of the lithium battery;

[0039] Discretization calculation formula for the state of charge SOC of the lithium battery:

[0040]

[0041] Among them, t represents time, △t is the interval sampling time, and C ba is the capacity of the lithium battery.

[0042] The standard quadratic form objective function is:

[0043]

[0044] In the formula:

[0045] u represents the vector to be optimized, which includes the output power P FC (k), P FC (k + 1), …, P FC (k + N p );

[0046] H is the Hessian matrix:

[0047] H = B T pB + q

[0048]

[0049] f is the linear term and can be expressed as the following formula:

[0050]

[0051] where f1 = 2(SOC(k) - SOC ref )diag[p], f2 = 2P FC (k)diag[q], and diag[] is a diagonal matrix.

[0052] After solving the standard quadratic form objective function, k + 1, k + 2, …, k + N p are obtained, arranged in sequence by time step. There are a total of N p fuel cell output power sequences. Select the (k + 1)-th one in the sequence as the predicted fuel cell output power at the next moment.

[0053] The method of directly allocating the power of the fuel cell and the lithium battery by using the state machine method includes:

[0054] For the low-power fast power response stage, the method of directly allocating the power of the fuel cell and the lithium battery by using the first state machine method includes:

[0055] P load = (0, 0.2P max )], P FC = 0.1P max , P bat = P load - P FC ;

[0056] P load = (0.2P max , 0.4P max )], P FC = 0.3P max , P bat= P load -P FC ;

[0057] P load = (0.4P max , 0.6P max , P FC = 0.5P max ,P bat = P load -P FC ;

[0058] P load = (0.6P max , 0.8P max , P FC = 0.7P max ,P bat = P load -P FC ;

[0059] P load = (0.8P max , 1.0P max , P FC = 0.9P max ,P bat = P load -P FC ;

[0060] Pl oa d = (1.0P max , 1.2P max , P FC = 1.1P max ,Pb at = Pl oa d - P FC ;

[0061] Wherein, P load is the required power of the vehicle, P FC , P bat are the output powers of the fuel cell and the lithium battery respectively, and P max represents the maximum required power of the vehicle;

[0062] For the high-power fast power response stage, the second state machine method is used to directly allocate the powers of the fuel cell and the lithium battery, including: on the basis of the first state machine method, for each required power P load interval, increase P FC by 0.05P max to enable the lithium battery to have higher energy storage potential to absorb the high-frequency power fluctuations of the vehicle.

[0063] For the low-power fast power response stage and the high-power fast power response stage, after directly allocating the power of the fuel cell and the lithium battery by using the corresponding state machine method, then according to the allocated P FC Obtain the hydrogen consumption of the fuel cell at the next moment through calculation by the fuel cell model, according to the allocated P bat Obtain the SOC of the lithium battery at the next moment through calculation by the lithium battery model.

[0064] The driving state of the vehicle is divided into different stages according to the power demand of the vehicle and the change of the power demand gradient, including:

[0065] Set the demand power threshold and the power demand gradient change threshold;

[0066] Compare the current demand power of the vehicle with the demand power threshold, and determine the driving state of the vehicle as the high-power stage or the low-power stage;

[0067] Compare the current power demand gradient change of the vehicle with the power demand gradient change threshold, determine the high-power stage as the high-power fast power response stage or the high-power slow power response stage, and determine the low-power stage as the low-power fast power response stage or the low-power slow power response stage.

[0068] The beneficial effects of the present invention are as follows:

[0069] The present invention combines the driving condition information of the vehicle to design the energy management of the fuel cell vehicle. According to the characteristics of the vehicle power demand under different working conditions, different methods are used for energy management, giving full play to the respective advantages of the state machine method and the model predictive control (MPC) method. Through the combination of the two, the global optimal energy distribution of the fuel cell vehicle is realized. Greatly reduce the hydrogen consumption of the proton exchange membrane fuel cell and maintain the stability of the lithium battery SOC within a certain range.

[0070] The present invention combines the optimization target and the calculation model through the MPC method to form a standard quadratic objective function, which is convenient for solving by using a general toolbox and has high calculation efficiency.

[0071] Other features and advantages of the present invention will be described in the subsequent specification or understood by implementing the present invention. Brief Description of the Drawings

[0072] Figure 1 It is a schematic flow chart within a prediction time step of the method of the embodiment of the present invention.

[0073] Figure 2 It is the vehicle power distribution result of the method of the embodiment of the present invention and the comparative method under the NEDC test cycle.

[0074] Figure 3 It is the vehicle power distribution result of the method of the embodiment of the present invention and the comparative method under the HWEFT test cycle.

[0075] Figure 4 It is the change trend of the SOC of the lithium battery of the vehicle under two test cycles for the method of the embodiment of the present invention and the comparative method. Specific Embodiments

[0076] The following describes the specific embodiments of the present invention with reference to the accompanying drawings.

[0077] An energy management method for a fuel cell vehicle in this embodiment, the fuel cell vehicle adopts a hybrid power system composed of a fuel cell and a lithium battery. The fuel cell is a proton exchange membrane fuel cell, the number of series-connected fuel cells is 455, the working voltage is 0 - 450V, the working current is 0 - 400A, the rated power is 100kW, the hydrogen and air supply pressures are 5 bar, and the working temperature is 293.15K - 333.15K. The open-circuit voltage of the lithium battery is 300V, and the rated capacity is 50Ah.

[0078] See Figure 1 , the energy management method includes:

[0079] S1. Establish a calculation model, the calculation model includes a fuel cell model, a lithium battery model and a vehicle dynamics model.

[0080] The establishment process of the fuel cell model in S11 includes:

[0081] Considering that the battery voltage of the proton exchange membrane fuel cell is generated by the electrochemical reaction of hydrogen and oxygen, establish the equation of the actual output circuit voltage V FC :

[0082] V FC = V Nernst - V act - V ohmic - V conc

[0083] In the formula, V Nernst , V act , V ohmic , V conc are the Nernst open-circuit voltage, activation polarization voltage, ohmic polarization voltage, and concentration polarization voltage respectively, and the expressions are respectively:

[0084]

[0085] V ohmic = I FC × R ohmic = I FC × (Rmem +R electrode ) = I FC ×R mem

[0086]

[0087] Wherein:

[0088] The above V Nernst 's expression is the numerical expression when the electrochemical reaction produces liquid water. T FC , T0 are the temperature of the fuel cell and the reference temperature (preferably 298.15 K) respectively, are the partial pressures of the anode and cathode reactants respectively.

[0089] The above V act 's expression is the theoretical expression according to the Butler-Volmer equation. The activation polarization voltage mainly comes from the energy required for electron transfer during the electrochemical reaction and the energy involved in breaking or forming chemical bonds on the anode and cathode sides. Therefore, there are activation losses on both the cathode and anode sides. Usually, the chemical reaction of oxygen is slower than the reduction reaction of hydrogen. ξ1, ξ2, ξ3, ξ4 are empirical coefficients, is the oxygen concentration at the cathode catalyst interface; I FC is the external circuit current.

[0090] The ohmic polarization voltage loss is composed of the electrode resistance and the internal resistance of the proton exchange membrane. Since the equivalent membrane impedance of the proton exchange membrane is the main factor of the ohmic polarization voltage, the ohmic polarization voltage loss can be described by the above V ohmic 's expression, R ohmic , R mem , R electrode are the total resistance, the proton exchange membrane resistance and the electrode resistance respectively.

[0091] The proton exchange membrane resistance depends to a large extent on the humidity of the proton exchange membrane, the current density and the operating temperature of the proton exchange membrane fuel cell. When the reactants are consumed, the potential difference generated due to the decrease in the reactant concentration generates a co-polarization voltage. This loss is the main reason for the rapid voltage drop of the fuel cell stack at high current densities. Therefore, it is expressed by the above V conc 's expression, I max is the maximum current of the fuel cell, R is the reaction constant, and F is the Faraday constant.

[0092] Establish a fuel cell output power model:

[0093] P FC = N cell V FC I FC

[0094] = N cell (V Nernst - V act - V ohmic - V conc )I FC

[0095] Among them, P FC is the output power of the fuel cell, and N cell is the number of fuel cells connected in series;

[0096] Establish a calculation formula for the theoretical hydrogen consumption rate:

[0097]

[0098] Among them, M H2 is the molar mass of hydrogen.

[0099] Through the fuel cell model of this embodiment, the output power and hydrogen consumption of the fuel cell can be calculated.

[0100] S12. The establishment of the lithium battery model includes:

[0101] Without considering the influence of temperature and degradation, the circuit of the lithium battery is simplified to an ideal resistor and an ideal open-circuit voltage source connected in series with the ideal resistor. The voltage value V OC of the ideal open-circuit voltage source is related to the SOC of the lithium battery:

[0102] V OC = a1SOC + a2SOC 2

[0103] Among them, a1 and a2 are constants obtained by fitting according to the actual operation data of the lithium battery;

[0104] According to the voltage formula of the lithium battery V ba = V OC - I ba R ba , the output power of the lithium battery is obtained:

[0105]

[0106] Among them, P bat , V ba , I ba are the output power of the lithium battery, the voltage of the lithium battery, and the external circuit current respectively, and R ba is the resistance value of the ideal resistor;

[0107] Obtain the discretization calculation formula for the state of charge SOC of the lithium battery:

[0108]

[0109] Among them, t represents time, △t is the interval sampling time, preferably 1 s, and C ba is the capacity of the lithium battery.

[0110] Through the lithium battery module of this embodiment, the SOC of the lithium battery can be calculated.

[0111] S13. Establish a vehicle dynamics model:

[0112] The vehicle dynamics model of this embodiment is the fuel cell vehicle dynamics model, which is a mathematical representation used to describe the motion and behavior of a vehicle in response to various forces, including engine power, braking, and external conditions such as road surface and gradient. It is usually used in simulation and control applications to predict the acceleration, speed, and handling characteristics of a vehicle under different operating conditions. The model expression is as follows:

[0113]

[0114] Among them, where P load 、P acc，r 、P roll，r 、P air，r are the required power of the vehicle (i.e., load power), resistance loss during acceleration, rolling resistance loss, and air resistance loss respectively, v ve is the driving speed of the vehicle, α roll 、α f 、α air are the acceleration coefficient, friction coefficient, and air resistance coefficient respectively, m and a are the vehicle mass and vehicle acceleration respectively, g is the acceleration due to gravity, A is the frontal area, and η t,e is the transmission efficiency of the vehicle motor.

[0115] Through the vehicle dynamics model, the required power P load of the vehicle can be calculated in real time.

[0116] In this embodiment, for the convenience of subsequent comparison and verification of the effectiveness of this solution with other comparison schemes, two standard driving cycle test data sets are directly selected: the New European Driving Cycle NEDC and the Highway Fuel Economy Test HWFET. The two data sets contain the speed and acceleration data of the vehicle during driving. The total mass of the vehicle in this embodiment is 3000 kg, the acceleration coefficient is 10, the rolling resistance coefficient is 0.03, the air resistance coefficient is 0.01, and the frontal area is 7.5 m 2 . Thus, the required power of the vehicle can be directly calculated according to the speed and acceleration data in the two test sets.

[0117] S2. Divide the current driving condition of the vehicle into different stages according to the power demand of the vehicle and the change in the power demand gradient, including a high-power fast power response stage, a high-power slow power response stage, a low-power fast power response stage, and a low-power slow power response stage. Specifically, it includes:

[0118] Set a demand power threshold of 0.5P max , and a power demand gradient change threshold δ;

[0119] Compare the current demand power P of the vehicle load with the demand power threshold of 0.5P max to determine whether the vehicle driving state is a high-power stage (High power, HP) or a low-power stage (Low power, LP);

[0120] Then further divide the two divided stages into a fast power response (Fast power, FP) and a slow power response stage (Slow power, SP). Specifically, it includes: Compare the current power demand gradient change of the vehicle with the power demand gradient change threshold δ, and determine the high-power stage as a high-power fast power response stage (HP-FP) or a high-power slow power response stage (HP-SP), and determine the low-power stage as a low-power fast power response stage (LP-SP) or a low-power slow power response stage (LP-FP).

[0121] The four different states divided are as follows:

[0122]

[0123] where P max is the maximum demand power of the vehicle, is the power gradient change of the vehicle; & represents "and".

[0124] In a typical vehicle driving cycle, the power demand curve can be divided into the above four intervals: LP-SP, LP-FP, HP-SP, and HP-FP. In the LP-SP stage, the vehicle operates at low power with relatively stable output, corresponding to the low-speed cruising stage. The LP-FP stage is characterized by low power and rapid change, usually occurring during the initial start-up stage or low-speed emergency braking. In the HP-SP stage, the vehicle operates in the high-speed cruising stage. The HP-FP stage corresponds to the end of vehicle start-up or the beginning of high-speed emergency braking.

[0125] Such as Figure 1As shown in the figure, in this embodiment, the driving state of the vehicle is first determined at which stage, and then a specific method is adopted for energy management for a specific stage. The State Machine (SM) method is used for power distribution in the FP stage, while the Model Predictive Control (MPC) method is used for the SP change stage. The MPC method is adopted in the SP stage to stabilize the SOC of the lithium battery and minimize the hydrogen consumption of the fuel cell, which can reduce the fuel cost. In the FP stage, the amplitude of the vehicle power changes greatly. In order to reduce the impact of frequent power fluctuations of the vehicle on the life of the fuel cell, a deterministic method such as the state machine is used to distribute the power of the fuel cell and the lithium battery, so that the power fluctuations of the vehicle are absorbed by the lithium battery.

[0126] S3. For the high-power slow power response stage and the low-power slow power response stage, adopt the MPC method to distribute the power of the fuel cell and the lithium battery, including:

[0127] S31. With the goal of a low degree of deviation of the state of charge of the lithium battery from the reference value and minimizing the hydrogen consumption of the fuel cell, establish an optimization target model:

[0128] minJ = J1 + J2

[0129] s.t. SOC min ≤ SOC(k + i|k) ≤ SOC max

[0130] 0 ≤ P FC (k + i|k) ≤ P FC,max

[0131] In the formula, the first target J1 is to reduce the degree of deviation of the SOC from the reference value during the actual operation process, which helps to slow down the phenomenon of accelerated aging of the lithium battery due to frequent charge and discharge; the first target J2 is for the hydrogen consumption index during the vehicle operation process, and the purpose is to reduce the vehicle operation cost. Specifically:

[0132]

[0133] SOC, SOC min 、SOC max are respectively the state of charge of the lithium battery and its minimum and maximum values, k represents the current moment, P FC 、P FC,max is the output power of the fuel cell and its maximum value; N p is the prediction interval step size, i represents the i-th time step within the N p time steps, p and q are weight coefficients, and SOC ref is the reference value of the state of charge of the lithium battery.

[0134] S32. Construct the optimized objective model and the calculation model established in step S1 into a standard quadratic objective function through the MPC method:

[0135]

[0136] In the formula:

[0137] u represents the vector to be optimized, which includes the output powers P FC (k), P FC (k + 1), …, P FC (k + N p ) at each moment;

[0138] H is the Hessian matrix:

[0139] H = B T pB + q

[0140]

[0141] f is the linear term and can be expressed as the following formula:

[0142]

[0143] Among them, f1 = 2(SOC(k) - SOC ref ), f2 = 2P FC (k)diag[q], p and q are the weight coefficients described above, and diag[] is a diagonal matrix.

[0144] S33. Based on the vehicle dynamics model in step S1, calculate the required power P load (k) at the current moment k;

[0145] Based on the fuel cell model, calculate the fuel cell output power P FC (k) at the current moment k;

[0146] Based on the lithium battery model, calculate the state of charge SOC(k) of the lithium battery at the current moment k;

[0147] Substitute P load (k), P FC (k), and SOC(k) into the standard quadratic objective function and use the existing toolbox to solve it to obtain the predicted fuel cell output power P FC (k + 1) and the lithium battery output power P bat (k + 1); It can be understood that in order to achieve the optimization goal, solving the objective function will obtain k + 1, k + 2, …, k + N arranged in time steps in sequence p a total of N pA fuel cell output power sequence, select the (k + 1)-th in the sequence as the predicted next moment fuel cell output power P FC (k + 1), and ignore the power corresponding to other time steps; then according to P FC (k + 1), calculate the hydrogen consumption of the fuel cell at the next moment through the fuel cell model, and according to P bat (k + 1), calculate the state of charge SOC(k + 1) of the lithium battery at the next moment through the lithium battery model;

[0148] Take P FC (k + 1), SOC(k + 1) as the input values of the new current moment, repeat S2, S3 for continuous loop prediction calculation to achieve power distribution;

[0149] S4. For the high-power fast power response stage and the low-power fast power response stage, use the state machine method to directly allocate the power of the fuel cell and the lithium battery, including:

[0150] For the low-power fast power response stage, use the first state machine method (SM1) to directly allocate the power of the fuel cell and the lithium battery, including:

[0151] Pl oa d=(0, 0.2P max , P FC =0.1P max , Pb at =Pl oa d - P FC ;

[0152] Pl oa d=(0.2P max , 0.4P max , P FC =0.3P max , Pb at =Pl oa d - P FC ;

[0153] Pl oa d=(0.4P max , 0.6P max , P FC =0.5P max , Pb at =Pl oa d - P FC ;

[0154] P load =(0.6P max , 0.8P max , P FC= 0.7P max , P bat = P load - P FC ;

[0155] P load = (0.8P max , 1.0P max , P FC = 0.9P max , P bat = P load - P FC ;

[0156] P load = (1.0P max , 1.2P max , P FC = 1.1P max , P bat = P load - P FC ;

[0157] Among them, P load is the required power of the vehicle, P FC , P bat are the output powers of the fuel cell and the lithium battery respectively, and P max represents the maximum required power of the vehicle;

[0158] For the high-power fast power response stage, the second state machine method (SM2) is used to directly allocate the powers of the fuel cell and the lithium battery, including:

[0159] On the basis of the first state machine method, for each required power P load interval, increase P FC by 0.05P max to enable the lithium battery to have higher energy storage potential to absorb the high-frequency power fluctuations of the vehicle.

[0160] It can be understood that after directly allocating the powers of the fuel cell and the lithium battery by using the corresponding state machine method, then according to the allocated P FC calculate the hydrogen consumption of the fuel cell at the next moment through the fuel cell model, and according to the allocated P bat calculate the SOC of the lithium battery at the next moment through the lithium battery model, and use the calculated value as the input value of the new current moment, and repeat S2 and S4 to achieve continuous cyclic prediction calculation.

[0161] To verify the effectiveness of the solution of this embodiment, under the condition that other condition parameters are the same, the separate SM (state machine) method and the separate MPC method are used for power distribution of the vehicle as comparative examples, and compared with the SM-MPC method of this embodiment to obtain the power distribution results of the vehicle and the change of the lithium battery SOC under two driving cycle test data. Among them, the separate SM method also includes the SM1 (first state machine) method and the SM2 (second state machine) method.

[0162] The power distribution results of the vehicle under the NEDC and HWEFT driving cycle test data are respectively as Figure 2 , Figure 3 shown. Figure 2 , Figure 3 In (a), (b), and (c) are the schematic diagrams of the results of the separate SM method, the separate MPC method, and the SM-MPC method respectively. In each figure, the abscissa is time, and the ordinate is P load . It can be seen from the figure that for the separate SM method, the fixed fuel cell power will cause excessive hydrogen consumption in most cases. Compared with the separate SM method, the separate MPC method performs better in tracking the power demand of the vehicle. The SM-MPC method of this embodiment is superior to the separate MPC method in tracking the power demand of the vehicle. This is because during the low-power response period, the MPC method stabilizes the output of the fuel cell, and the SM method is used during the high-power response period, thereby reducing the ability of the lithium-ion battery to absorb high-frequency power fluctuations. In other words, it prolongs the life of the fuel cell stack.

[0163] The corresponding SOC change curves of the power distribution results of the three methods under the two driving cycle test data are as Figure 4 shown. Figure 4 In (a) and (b) are the schematic diagrams of the results under the NEDC and HWEFT driving cycle test data respectively. In the figure, the abscissa is time, and the ordinate is SOC. It can be seen from the figure that under the condition of the NEDC driving cycle test data, the SOC assigned by the separate SM method shows a gradually increasing trend. This is mainly because this situation is mostly in the low-power demand stage, and the SM1 method can continuously charge the lithium battery, resulting in its SOC being higher than the reference point. In contrast, the separate MPC method shows an SOC value lower than the reference point 0.6 most of the time. And the SM-MPC method of this embodiment can keep the SOC near the reference value. In the HWEFT driving cycle test, it can still be observed that the SM-MPC method more effectively maintains the SOC value of the lithium battery.

[0164] In terms of energy consumption, for the NEDC test cycle, the hydrogen consumption of the three methods at 1180 seconds is different. The SM method alone consumes 286.85 g of hydrogen, the MPC method alone consumes 275.74 g of hydrogen, and the SM-MPC method consumes 268.63 g of hydrogen. The SM-MPC method has the lowest hydrogen consumption, which is 7.11 grams lower than the MPC method and 18.22 grams lower than the SM method. For the 1500-second HWEFT test, the SM method alone consumes 705.23 g of hydrogen, the MPC method alone consumes 696.31 g of hydrogen, and the SM-MPC method consumes 694.42 g of hydrogen. The SM-MPC method uses the least hydrogen, with a hydrogen consumption reduction of 1.89 grams compared to the MPC method alone and 10.18 grams compared to the SM method alone.

[0165] In summary, this embodiment combines the driving condition information of the vehicle to design the energy management of the fuel cell vehicle, significantly reducing the hydrogen consumption of the proton exchange membrane fuel cell and maintaining the stability of the SOC of the lithium battery within a certain range.

[0166] Those of ordinary skill in the art can understand that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, for those skilled in the art, they can still modify the technical solutions described in the foregoing embodiments or perform equivalent replacements for some of the technical features. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An energy management method for a fuel cell vehicle, wherein the fuel cell vehicle adopts a hybrid power system composed of a fuel cell and a lithium battery, characterized in that: The energy management method comprises: Establishing a calculation model, wherein the calculation model includes a vehicle dynamics model, a fuel cell model and a lithium battery model; The current driving condition of the vehicle is divided into different stages according to the power demand of the vehicle and the gradient change of the power demand, including a high-power fast power response stage, a high-power slow power response stage, a low-power fast power response stage and a low-power slow power response stage; For the high-power slow power response stage and the low-power slow power response stage, the MPC method is used to distribute the power of the fuel cell and the lithium battery, including: Establishing an optimization target model with the goal of reducing the degree to which the state of charge of the lithium battery deviates from a reference value and minimizing the hydrogen consumption of the fuel cell; The optimization target model and the calculation model are constructed into a standard quadratic objective function by using an MPC method; Based on the dynamic model of the vehicle, calculating and obtaining the required power at the current moment; Based on the fuel cell model, calculating and obtaining the fuel cell output power at the current moment; Based on the lithium battery model, calculating and obtaining the current state of charge of the lithium battery; Substituting the current power demand, fuel cell output power and lithium battery state of charge into the standard quadratic objective function for solution, obtaining the predicted fuel cell output power and lithium battery output power at the next moment, and then obtaining the fuel cell hydrogen consumption and lithium battery state of charge at the next moment according to the calculation model, thereby realizing power allocation; For the high-power fast power response stage and the low-power fast power response stage, a state machine method is used to directly allocate the power of the fuel cell and the lithium battery.

2. The energy management method for a fuel cell vehicle according to claim 1, characterized in that: The optimization target model is: minJ=J1+J2 s.t.SOC min ≤SOC(k+i|k)≤SOC max 0≤P FC (k+i|k)≤P FC,max in, In the formula, SOC, SOC min , SOC max are the state of charge of the lithium battery and its minimum and maximum values, k represents the current moment, P FC , P FC,max is the fuel cell output power and its maximum value; N p is the prediction interval step size, i represents the p The i-th time step within the time step, p, q are weight coefficients, SOC ref It is the reference value of the charge state of lithium battery.

3. The energy management method for a fuel cell vehicle according to claim 2, characterized in that: The vehicle dynamics model is: Among them, P load , P acc，r , P roll，r , P air，r are the vehicle's required power, drag loss during acceleration, rolling resistance loss, and air resistance loss, respectively. ve is the vehicle’s speed, α roll , α f , α air They are acceleration coefficient, friction coefficient and air resistance coefficient, m and a are vehicle mass and vehicle acceleration, g is gravity acceleration, A is windward area, η t,e is the vehicle motor transmission efficiency.

4. The energy management method for a fuel cell vehicle according to claim 3, characterized in that: The fuel cell model includes: P FC =N cell V FC I FC =N cell (V Nernst -V act -V ohmic -V conc )I FC Among them, N cell 、V FC ,I FC are the number of fuel cells in series, the actual output circuit voltage, and the external circuit current; V Nernst 、V act 、V ohmic 、V conc They are Nernst open circuit voltage, activation polarization voltage, ohmic polarization voltage, and concentration polarization voltage; Theoretical hydrogen consumption rate calculation formula: Among them, M H2 is the molar mass of hydrogen, and F is the Faraday constant.

5. The energy management method for a fuel cell vehicle according to claim 4, characterized in that: The lithium battery model includes: Lithium battery output power: The circuit of the lithium battery is simplified to an ideal resistor and an ideal open-circuit voltage source connected in series with the ideal resistor. bat 、V ba ,I ba are the lithium battery output power, lithium battery voltage and external circuit current, R ba 、V OC are respectively the resistance value of the ideal resistor and the voltage value of the ideal open-circuit voltage source; Among them, V OC =a1SOC+a2SOC 2 , a1, a2 are constants obtained by fitting the actual operation data of lithium batteries; The discrete calculation formula of the state of charge SOC of a lithium battery is: Among them, t represents time, △t is the interval sampling time, C ba is the capacity of the lithium battery.

6. The energy management method for a fuel cell vehicle according to claim 5, characterized in that: The standard quadratic objective function is: Where: u represents the vector to be optimized, which includes the output power P at each moment FC (k), P FC (k+1),…,P FC (k+N p ); H is the Hessian matrix: H=B T pB+q f is a linear term, which can be expressed as follows: Where f1 = 2(SOC(k)-SOC ref )diag[p], f2=2P FC (k)diag[q], diag[] is a diagonal matrix.

7. The energy management method for a fuel cell vehicle according to claim 2, characterized in that: After solving the standard quadratic objective function, k+1, k+2, ..., k+N are obtained in sequence according to the time step. p Total N p A fuel cell output power sequence is selected, and the k+1th one in the sequence is selected as the predicted fuel cell output power at the next moment.

8. The energy management method for a fuel cell vehicle according to claim 1, characterized in that: The state machine method is used to directly distribute the power of the fuel cell and the lithium battery, including: In the low-power fast power response stage, a first state machine method is used to directly allocate the power of the fuel cell and the lithium battery, including: P load =(0,0.2P max ],P FC =0.1P max ,P bat =P load -P FC ; P load =(0.2P max ,0.4P max ],P FC =0.3P max ,P bat =P load -P FC ; P load =(0.4P max ,0.6P max ],P FC =0.5P max ,P bat =P load -P FC ; P load =(0.6P max ,0.8P max ],P FC =0.7P max ,P bat =P load -P FC ; P load =(0.8P max ,1.0P max ],P FC =0.9P max ,P bat =P load -P FC ; P load =(1.0P max ,1.2P max ],P FC =1.1P max ,P bat =P load -P FC ; Among them, P load is the required power of the vehicle, P FC , P bat are the output power of fuel cell and lithium battery respectively, P max Indicates the maximum required power of the vehicle; In the high-power fast power response stage, the second state machine method is used to directly allocate the power of the fuel cell and the lithium battery, including: on the basis of the first state machine method, for each required power P load P in the interval FC Increase 0.05P max , giving lithium batteries a higher energy storage potential to absorb the vehicle's high-frequency power fluctuations.

9. The energy management method for a fuel cell vehicle according to claim 8, characterized in that: For the low-power fast power response stage and the high-power fast power response stage, the power of the fuel cell and the lithium battery is directly allocated using the corresponding state machine method, and then the P FC The fuel cell hydrogen consumption at the next moment is obtained by calculating the fuel cell model, and the P bat The lithium battery SOC at the next moment is obtained by calculating the lithium battery model.

10. The energy management method for a fuel cell vehicle according to claim 1, characterized in that: The vehicle driving state is divided into different stages according to the power demand of the vehicle and the gradient change of the power demand, including: Set the required power threshold and power requirement gradient change threshold; Comparing the current required power of the vehicle with the required power threshold, and determining the vehicle driving state as a high power stage or a low power stage; The vehicle's current power demand gradient change is compared with the power demand gradient change threshold, and the high power stage is determined as a high power fast power response stage or a high power slow power response stage, and the low power stage is determined as a low power fast power response stage or a low power slow power response stage.

Images