A method and system for energy management of multiple stacks of fuel cells based on fuzzy control

Abstract

The application discloses a kind of energy management method and system of multi-stack fuel cell based on fuzzy control, the method includes the following steps: constructing the energy system including multi-stack fuel cell and lithium battery;The demand power of energy system and the state of charge of lithium battery are as input variable, the output power of multi-stack fuel cell is as output variable, and fuzzy logic controller is constructed;According to the relationship between the output power of multi-stack fuel cell and lithium battery and the hydrogen consumption of energy system, the hydrogen consumption function of energy system is established, and the optimization function is established to minimize the value of hydrogen consumption function;Optimization function is solved, and the optimal parameter of fuzzy logic controller is obtained, and the output power of multi-stack fuel cell is controlled according to optimal parameter.The method of the application optimizes the energy consumption of fuel cell, reduces the use of hydrogen fuel, and improves the energy utilization efficiency of the system.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of energy management, in particular to a multi-stack fuel cell energy management method and system based on fuzzy control. BACKGROUND

[0002] Fuel cell stacks, as a clean and efficient energy conversion technology, have shown their potential in providing grid stability, achieving load balancing, supporting distributed generation, and integrating with renewable energy sources. These technical advantages make fuel cell stacks play an important role in the clean transformation of grid energy.

[0003] In order to improve the performance and efficiency of hydrogen fuel cell stacks, existing control technologies usually use fuzzy logic controllers. Fuzzy logic controllers define input and output variables, establish fuzzy sets and membership functions, develop fuzzy rule bases, perform fuzzy processing, fuzzy reasoning and defuzzification, and finally achieve precise control of fuel cell stacks. This method can quickly respond to system changes, adapt to complex and variable working conditions, improve the robustness and adaptability of the system, and is particularly suitable for dynamic changes and multi-variable fuel cell systems.

[0004] Although fuzzy logic controllers have shown certain advantages in fuel cell stack control, there are still some problems to be solved. The design and parameter adjustment of fuzzy logic controllers often depend on expert experience and trial-and-error methods, which may lead to inaccurate and optimized controller design. And the existing fuzzy logic controller may have deficiencies in real-time performance and computational efficiency, especially in high dynamic response application scenarios. SUMMARY

[0005] The present application proposes a multi-stack fuel cell energy management method and system based on fuzzy control, which solves the problem of poor real-time performance and computational efficiency of existing fuel cell stack control methods.

[0006] To solve the above technical problems, the present application provides a multi-stack fuel cell energy management method based on fuzzy control, comprising the following steps:

[0007] Step S1: Construct an energy system including multi-stack fuel cells and lithium batteries;

[0008] Step S2: Take the demand power of the energy system and the state of charge of the lithium battery as input variables, and take the output power of the multi-stack fuel cell as an output variable, and construct a fuzzy logic controller;

[0009] Step S3: Establish a hydrogen consumption function of the energy system according to the relationship between the output power of the multi-stack fuel cell and the lithium battery and the hydrogen consumption, and establish an optimization function with the goal of minimizing the value of the hydrogen consumption function.

[0010] Step S4: solving the optimization function to obtain the optimal parameters of the fuzzy logic controller, and controlling the output power of the multiple fuel cells according to the optimal parameters.

[0011] Preferably, the constructing the fuzzy logic controller in step S2 comprises the following steps:

[0012] Step S21: defining a plurality of membership functions, and deriving fuzzy sets of the input variables and output variables according to the plurality of membership functions;

[0013] Step S22: defining a mapping relationship between the fuzzy sets of the input variables and output variables as fuzzy rules;

[0014] Step S23: defining free variables of the plurality of membership functions, and taking the free variables as parameters of the fuzzy logic controller.

[0015] Preferably, seven membership functions mf1-mf7 are constructed in step S21, and mf1-mf7 correspond to very low, medium low, low, medium, high, medium high and very high respectively.

[0016] Preferably, the fuzzy rules in step S22 comprise:

[0017] (1) when the state of charge of the lithium battery is low and the power demand of the energy system is also low, the fuel cell should output high power;

[0018] (2) when the state of charge of the lithium battery is low and the power demand of the energy system is high, the fuel cell should output high power;

[0019] (3) when the state of charge of the lithium battery is high and the power demand of the energy system is low, the fuel cell should output low power;

[0020] (4) when the state of charge of the lithium battery is high and the power demand of the energy system is also high, the fuel cell should output low power.

[0021] Preferably, the free variables in step S23 comprise a vertex x1, a lower left corner x2, a lower right corner x3, a lower scale x4, a left lower hysteresis x5 and a right lower hysteresis x6.

[0022] Preferably, the expression of the output power of the multiple fuel cells in step S3 is:

[0023] P FC =V FC I FC =nV cell I FC ;

[0024] V FC = nV cell ;

[0025] V cell = E Nerest -V act -V ohm -V con ;

[0026]

[0027] V ohm = R FC I FC ;

[0028]

[0029] In the above formula, P FC is the output power of the multi-stack fuel cell; V FC is the output voltage of the multi-stack fuel cell; I FC is the output current of the multi-stack fuel cell; n is the number of fuel cells in the multi-stack fuel cell; V cell , E Nerest , V act , V ohm , V con , T FC , R FC , k con and I FC max are the output voltage, Nernst voltage, activation polarization voltage drop, ohmic voltage drop, concentration polarization voltage drop, working temperature, hydrogen pressure, oxygen pressure, oxygen concentration, internal resistance, concentration polarization voltage coefficient and maximum current of a single fuel cell, respectively; ε1, ε2, ε3 and ε4 are constants, respectively.

[0030] Preferably, the expression of the hydrogen consumption function in step S3 is:

[0031] H FCT (X) = H FC (X) + H bat (X);

[0032] X = [x1 x2 x3 x4 x5 x6] T ;

[0033] H FC (X) = αX

[0034] α = [α1α2α3α4α5α6]

[0035] H bat (X) = βX;

[0036] β = [β1β2β3β4β5β6]

[0037]

[0038] In the above formula, H FC (X), H bat (X) are equivalent hydrogen consumption functions of the fuel cell and the lithium battery, respectively; a, b are weight matrices; a1, a2, a3, a4, a5, a6, b1, b2, b3, b4, b5, b6 are weight coefficients; H FC , H bat represent the hydrogen consumption of the multi-stack fuel cell and the lithium battery, respectively; P FC is the output power of the multi-stack fuel cell; η FC is the efficiency of the multi-stack fuel cell; P bat is the output power of the lithium battery; η bat is the efficiency of the lithium battery; η c , η d are the charging efficiency and the discharging efficiency of the lithium battery, respectively; t ini and t end represent the initial time and the end time, respectively.

[0039] Preferably, the optimization function is solved by using a surrogate model optimization algorithm and an adaptive particle swarm optimization algorithm in step S4, including the following steps:

[0040] Step S41: initialize the parameter set of the fuzzy logic controller and the parameters of the surrogate model optimization algorithm;

[0041] Step S42: calculate the hydrogen consumption corresponding to each group of parameters in the parameter set, and take the group of parameters with the minimum hydrogen consumption as the first optimal solution;

[0042] Step S43: construct a surrogate model according to the current parameter set, use the surrogate model to interpolate the optimization function to obtain a surrogate function, randomly generate a candidate solution set of the surrogate function, and select a second optimal solution from the candidate solution set;

[0043] Step S44: compare the hydrogen consumptions corresponding to the first optimal solution and the second optimal solution, take the solution with the smaller hydrogen consumption as the current optimal solution, and add the current optimal solution to the parameter set;

[0044] Step S45: repeat steps S42 to S44 until the current iteration number reaches the set maximum iteration number, and output the parameter set;

[0045] Step S46: input the parameter set into the adaptive particle swarm optimization algorithm to obtain an optimal parameter combination.

[0046] Preferably, the step S43 of selecting a second optimal solution from the candidate solution set comprises the following steps:

[0047] Step S431: calculate the shortest distance between each candidate point and the sampling point set in the candidate solution set

[0048]

[0049] wherein, represents the position of the point in the training set;

[0050] Step S432: set the maximum and minimum values of the agent function value S max , S min , the maximum and minimum values of the shortest distance between the candidate point and the sampling point D max , D min are:

[0051]

[0052]

[0053] Step S433: generate an ascending weight set {σ1,σ2,…,σ κ} between 0 and 1, if mod(τ,κ)≠0, let σ S ←σ mod(τ,κ) , otherwise let σ S ←σ κ , wherein mod() is the modulus function;

[0054] Step S434: let σ D ←1-σ S ;

[0055] Step S435: if S min ≠S max , let otherwise let

[0056] Step S436: if D min ≠D max , let otherwise let

[0057] Step S437: update the current optimal solution:

[0058]

[0059] Step S438: repeating steps S435 to S437 until the current iteration number reaches the set maximum iteration number, and outputting the current optimal solution as a second optimal solution.

[0060] The application also provides a multi-stack fuel cell energy management system based on fuzzy control, which is realized based on the multi-stack fuel cell energy management method based on fuzzy control.

[0061] The data acquisition module is configured to acquire operation data of the multi-stack fuel cell and the lithium battery.

[0062] The energy management module is configured to distribute output power of the multi-stack fuel cell and the lithium battery according to the operation data acquired by the data acquisition module, so as to meet the demand power of the system.

[0063] The control module is configured to control the output power of the multi-stack fuel cell and the lithium battery according to the control signal output by the energy management module.

[0064] The application has at least the following advantages:

[0065] 1. The fuzzy logic controller can process uncertainty and nonlinearity, and can realize accurate control of the output power of the multi-stack fuel cell by taking the demand power and the state of charge of the lithium battery as input variables.

[0066] 2. By accurately controlling the output power of the multi-stack fuel cell, the energy consumption of the fuel cell is optimized, the use of hydrogen fuel is reduced, and the energy utilization efficiency of the system is improved. BRIEF DESCRIPTION OF DRAWINGS

[0067] Figure 1 The method flowchart of the embodiment of the application is shown in the figure.

[0068] Figure 2 The system framework diagram of the embodiment of the application is shown in the figure.

[0069] Figure 3 The fuzzy logic controller and membership function of the embodiment of the application are shown in the figure.

[0070] Figure 4 The relationship between the parameters of the fuzzy logic controller and the hydrogen consumption of the embodiment of the application is shown in the figure.

[0071] Figure 5 The flowchart of the proxy model optimization algorithm of the embodiment of the application is shown in the figure.

[0072] Figure 6A flowchart of an adaptive particle swarm optimization algorithm of an embodiment of the present application. DETAILED DESCRIPTION

[0073] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the protection scope of the present application.

[0074] As shown in Figure 1 , the embodiment of the present application provides an energy management method for multiple fuel cell stacks based on fuzzy control, comprising the following steps:

[0075] Step S1: Construct an energy system comprising multiple fuel cell stacks and lithium batteries.

[0076] As shown in Figure 2 , the system framework of the embodiment of the present application consists of two main parts: a physical system and a cloud system, including an entity energy system and a fuel cell stack cloud model.

[0077] Specifically, the entity energy system contains a multiple fuel cell stack power generation system, a lithium battery energy storage system, a data acquisition system, a local controller and a DC / DC converter.

[0078] The fuel cell stack cloud model contains a fuel cell stack digital twin model and a parameter control system. The fuel cell stack cloud model is a model constructed in the cloud based on cloud computing technology, which can optimize the calculation of the controller parameters in the cloud combined with the fuel cell stack digital twin model, and send them to the entity energy system, control the entity energy system with the local controller, and obtain the optimal and hydrogen-saving operation state of the entity energy system.

[0079] Specifically, a fuel cell is a device that generates electrical energy through an electrochemical reaction between hydrogen and oxygen, and the output voltage of a single fuel cell is:

[0080] V cell =E Nerest -V act -V ohm -V con ;

[0081]

[0082] V ohm =R FC I FC ;

[0083]

[0084] In the above formula, V cell , E Nerest , V act , V ohm and V con represent the output voltage, the Nernst voltage, the activation polarization voltage drop, the ohmic voltage drop and the concentration polarization voltage drop of a single fuel cell respectively; T FC , and represent the working temperature, the hydrogen pressure and the oxygen pressure of the fuel cell respectively; ε1, ε2, ε3 and ε4 are constants respectively; is the oxygen concentration; R FC , I FC , k con and I FC max are the internal resistance of the fuel cell, the current of the fuel cell, the concentration polarization voltage coefficient and the maximum current of the fuel cell respectively.

[0085] In practical applications, the fuel cell stack is composed of a plurality of single cells connected in series, and the output voltage of the multi-stack fuel cell is:

[0086] V FC = nV cell .

[0087] In the formula, n is the number of single fuel cells connected in series in the multi-stack fuel cell.

[0088] Therefore, the output power of the multi-stack fuel cell is:

[0089] P FC = V FC I FC = nV cell I FC = n(E Nerest -V act -V ohm -V con )I FC .

[0090] Step S2: Taking the demand power of the physical energy system and the state of charge of the lithium battery as input variables and taking the output power of the multi-stack fuel cell as an output variable, a fuzzy logic controller is constructed.

[0091] Specifically, the embodiment of the present application adopts the second type of fuzzy logic control of the Mamdani fuzzy system to process uncertainty, and the typical structure is as shown in Figure 3As shown, the input and output variables are described by seven membership functions (mf1-mf7). Assume that the seven membership functions mf1 to mf7 here describe the degree of a variable as "very small", "moderately small", "slightly small", "moderate", "larger", "moderately large" and "very large", respectively.

[0092] The fuzzy rules of the fuzzy logic controller are defined by the following rules:

[0093] (1) When the state of charge of the lithium battery is low and the power demand of the energy system is also low, the fuel cell should output a higher power;

[0094] (2) When the state of charge of the lithium battery is low and the power demand of the energy system is high, the fuel cell should output a higher power;

[0095] (3) When the state of charge of the lithium battery is high and the power demand of the energy system is low, the fuel cell should output a lower power;

[0096] (4) When the state of charge of the lithium battery is high and the power demand of the energy system is also high, the fuel cell should output a lower power.

[0097] The final fuzzy rules are shown in Table 1.

[0098] Table 1 Fuzzy rules of the fuzzy logic controller

[0099]

[0100]

[0101] Each membership function has six free variables, namely the vertex, the lower left corner, the lower right corner, the lower scale, the lower left hysteresis and the lower right hysteresis, and the six parameters of each membership function are denoted as x1, x2, x3, x4, x5 and x6. By optimizing these parameters, the purpose of optimizing the fuzzy logic controller can be achieved, so as to realize the low hydrogen consumption of the fuel cell stack. In the embodiment of the present application, the input variable of the fuzzy logic controller is defined as the demand power of the physical energy system, and the output variable is the output power of the multi-stack fuel cell.

[0102] The six parameters of each membership function, namely the vertex x1, the lower left corner x2, the lower right corner x3, the lower scale x4, the lower left hysteresis x5 and the lower right hysteresis x6, are used as optimization parameters, and these parameters together constitute the decision vector X. The relationship between the hydrogen consumption of the multi-stack fuel cell and the parameters of the fuzzy logic controller is shown in Figure 4 As shown, by modifying the parameters of the fuzzy logic controller, the performance of the controlled system can be changed, thereby changing the hydrogen consumption of the multi-stack fuel cell. In the embodiment of the present application, the hydrogen consumption H FCT(X) is composed of two parts: one part is the actual hydrogen consumption of the multi-stack fuel cell, and the other part is the equivalent hydrogen consumption of the multi-stack fuel cell:

[0103] H FCT (X)=H FC (X)+H bat (X);

[0104]

[0105]

[0106] X=[x1 x2 x3 x4 x5 x6] T

[0107] In the above formula, H FC (X), H bat (X) are the hydrogen consumption function of the fuel cell and the equivalent hydrogen consumption function of the fuel cell, respectively; X represents the decision vector; α and β represent the weight matrix, respectively; α1, α2, α3, α4, α5, α6, β1, β2, β3, β4, β5, and β6 represent the weight coefficients.

[0108] Step S3: Establishing the hydrogen consumption function of the energy system according to the relationship between the output power and the hydrogen consumption of the multi-stack fuel cell and the lithium battery, and establishing an optimization function with the goal of minimizing the value of the hydrogen consumption function.

[0109] Specifically, in combination with the objective function and the decision variable, the optimization problem of the optimal hydrogen consumption of the fuel cell stack is obtained as follows:

[0110]

[0111] Wherein, X=[x1,x2,…,x n ] represents the decision vector, that is, the fuzzy controller parameter; n is the dimension of the decision vector; and are the upper limit and the lower limit.

[0112] Step S4: Solving the optimization function to obtain the optimal parameters of the fuzzy logic controller, and controlling the output power of the multi-stack fuel cell according to the optimal parameters.

[0113] Since the objective function of the embodiment of the application does not have an explicit function expression, it has black box characteristics, and therefore the gradient information of the objective function is unknown and cannot be solved by some traditional mathematical programming algorithms. Considering the characteristics of the black box function, the embodiment of the application selects an algorithm combining the surrogate model optimization algorithm (SO) and the adaptive particle swarm optimization algorithm (APSO) to solve the objective function.

[0114] The SO algorithm has the advantages that it can be optimized based on an initial training set, generate a surrogate model using an interpolation function, and then generate multiple candidate solutions near the current optimal feasible solution at each iteration step. In addition, according to the surrogate standard and the distance standard, the best candidate solution is selected, and the training set, the current best objective value and the best feasible solution are updated. However, this algorithm often falls into the dilemma of inefficient search in the later stage. Therefore, in the embodiment of the present application, the SO algorithm and the APSO algorithm are combined, the SO algorithm is used to determine the best feasible solution in the early stage, and the APSO algorithm is used to further refine the feasible solution in the later stage, so that the hydrogen consumption of the whole model is minimized, and the numerical value is more accurate.

[0115] The SO algorithm is a commonly used method in the fields of engineering calculation and machine learning, which is used to solve optimization problems of complex functions. This method constructs a relatively simple new function, called a surrogate model, by collecting feature points, and then optimizes this model to obtain the optimal solution. The surrogate model is usually divided into interpolation type and fitting type, which can effectively approximate the original complex function, and has a significant advantage especially when the function calculation cost is high or the calculation speed is slow.

[0116] The solving process of the SO algorithm adopted in the embodiment of the present application is as shown in Figure 4 The steps include the following steps:

[0117] Step S41: initialize the parameter set of the fuzzy logic controller and the parameters of the surrogate model optimization algorithm.

[0118] Step S42: calculate the hydrogen consumption corresponding to each group of parameters in the parameter set, and take the group of parameters with the minimum hydrogen consumption as the first optimal solution X opt :

[0119]

[0120] In the above formula, T represents the constructed training set,

[0121] Step S43: construct a surrogate model according to the current parameter set, use the surrogate model to interpolate the optimization function to obtain a surrogate function, randomly generate a candidate solution set of the surrogate function, and select a second optimal solution from the candidate solution set.

[0122] Specifically, the surrogate model adopted in the embodiment of the present application is an RBF interpolation model, and its expression is:

[0123]

[0124] In the formula, ‖·‖ represents the Euclidean norm; N represents the number of samples in the training set; p represents the dimension of ; b represents a constant; is the current decision variable; is the ith decision variable; a i denotes a coefficient, a i , ρ and b can be determined by the following equation:

[0125]

[0126] where Ψ, Γ1, Γ2, E are only operator symbols and have no special meaning, and are respectively:

[0127]

[0128] Here, Φ is a matrix composed of .

[0129] The candidate solution set of the agent function is randomly generated, including the following steps:

[0130] Step S4301: input the perturbation step size d and the current optimal solution

[0131] Step S4302: let where is a dimensional column vector, and the ith candidate point perturbation count x flag ←0;

[0132] Step S4303: determine whether the jth element in r per at this time [r per ] j is less than the set probability u per , if so, execute step S4304, otherwise execute step S4305;

[0133] Step S4304: let where normrnd() is a normal random number, and x flag ←x flag +1;

[0134] Step S4305: repeat steps S4303 to S4304 until the determination is completed for all ;

[0135] Step S4306: if x flag ==0, randomly select an integer between 1 and and assign it to j;

[0136] Step S4307: update

[0137] Step S4308: Repeat steps S4302 to S4307 until the current iteration number reaches the set maximum iteration number.

[0138] Selecting the second optimal solution from the candidate solution set comprises the following steps:

[0139] Step S4311: Let the candidate point set generated in the foregoing steps be denoted as Let the sampling point set be denoted as Calculate the shortest distance between each candidate point and the sampling point set in the candidate solution set.

[0140]

[0141] In the foregoing formula, represents the position of a point in the training set.

[0142] Step S4312: Set the maximum and minimum values of the agent function value S max , S min , the maximum and minimum values of the shortest distance between the candidate point and the sampling point D max , D min , as follows:

[0143]

[0144] Step S4313: Generate an ascending weight set {σ1, σ2, …, σ κ} between 0 and 1, and if mod(τ, κ)≠0, let σ S ←σ mod(τ,κ) , otherwise let σ S ←σ κ , where mod() is the modulus function.

[0145] Step S4314: Let σ D ←1-σ S .

[0146] Step S4315: If S min ≠S max , let , otherwise let

[0147] Step S4316: If D min ≠D max , let , otherwise let

[0148] Step S4317: Update the current optimal solution:

[0149]

[0150] Step S4318: repeating step S435 to step S437 until the current iteration number reaches the set maximum iteration number, outputting the current optimal solution as the second optimal solution

[0151] Step S44: comparing the hydrogen consumption amounts corresponding to the first optimal solution and the second optimal solution, taking the solution with the smaller hydrogen consumption amount as the current optimal solution, and adding the current optimal solution to the parameter set.

[0152] Specifically, if the current optimal solution, the current success and failure number are updated as:

[0153]

[0154] At this time, if S success exceeds the set maximum success number the current perturbation step size χ n+1 = 2χ n , the current success number S success = 0.

[0155] If the current success and failure number are updated as:

[0156]

[0157] At this time, if F fail exceeds the set maximum failure number the current perturbation step size χ n+1 = 0.5χ n , the current failure number F fail = 0.

[0158] Step S45: repeating step S42 to step S44 until the current iteration number reaches the set maximum iteration number, outputting the parameter set;

[0159] Step S46: inputting the parameter set into the adaptive particle swarm optimization algorithm to obtain the optimal parameter combination.

[0160] The solution process of the adaptive particle swarm optimization algorithm (APSO) adopted in the embodiments of the present application is as follows: Figure 5As shown. The adaptive particle swarm optimization algorithm (APSO) is an improved algorithm based on the traditional particle swarm optimization (PSO). It improves the performance and convergence speed of the algorithm by adaptively adjusting the algorithm parameters. The core of APSO is to dynamically adjust the algorithm parameters, such as inertia weight and learning factor, according to the convergence of the group, so that particles can search the solution space more effectively. In an embodiment of the present invention, the feasible solution obtained by the above-mentioned agent model optimization algorithm (SO) can be further refined by the APSO algorithm to obtain the optimal solution.

[0161] First, input the feasible solution set obtained by the SO algorithm above, set k = 1, and define the maximum number of iterations K APSO , then initialize the particle swarm and define the initial and final inertia weights Initial and final teach-in coefficients Initial and final population learning coefficients Initialize the historical optimal position for each individual and hydrogen consumption The entire group initializes the historical optimal position X pop and hydrogen consumption F FCT (X pop ).

[0162] When k≤K APSO When , loop k from 1 to N:

[0163]

[0164] And update according to the formula X pop ,F FCT (X pop ), make k=k+1, until the loop condition is no longer met, end the loop, and take the control parameters corresponding to the optimal individual as the optimal control parameter combination.

[0165] The method of the present invention realizes the economic optimization of the fuel cell stack supporting the power grid by constructing a digital twin model of the fuel cell stack, combining it with fuzzy logic control based on uncertain rules, and utilizing an algorithm based on a substitute model and an adaptive particle swarm optimization algorithm. Through this method, the hydrogen consumption of the energy system can be reduced, energy efficiency can be improved, operating costs can be reduced, the service life of the fuel cell stack can be extended, and its competitiveness in practical applications can be enhanced. The present invention not only has direct significance for improving the performance and economic benefits of the fuel cell stack, but also provides important data support and technical guidance for the development of the entire hydrogen energy industry chain, and promotes the widespread application of fuel cell technology in the power grid field.

[0166] The embodiment of the present application also provides a fuzzy control-based energy management system of multiple fuel cell stacks, which is realized based on the fuzzy control-based energy management method of multiple fuel cell stacks.

[0167] The data acquisition module is used for acquiring the operation data of the multiple fuel cell stacks and the lithium battery.

[0168] The energy management module is used for distributing the output power of the multiple fuel cell stacks and the lithium battery according to the operation data acquired by the data acquisition module, so as to meet the demand power of the system.

[0169] The control module is used for controlling the output power of the multiple fuel cell stacks and the lithium battery according to the control signal output by the energy management module.

[0170] The technical features of the above embodiments can be combined arbitrarily, and in order to make the description simple, all possible combinations of the technical features in the above embodiments are not described, and only the preferred embodiments of the present application are expressed, which are described in detail, but it should not be understood as the limitation of the patent scope of the present application. As long as the combinations of the technical features do not exist contradictory, they should be considered as the range recorded in the specification.

[0171] It should be pointed out that, for those skilled in the art, without departing from the concept of the present application, a number of modifications and improvements can be made, which all belong to the protection scope of the present application. Therefore, the protection scope of the present application patent should be subject to the appended claims.

Claims

1. A method for energy management of multiple fuel cell stacks based on fuzzy control, characterized in that: The following steps are involved: Step S1: constructing an energy system including multiple fuel cell stacks and lithium batteries; Step S2: using the required power of the energy system and the state of charge of the lithium battery as input variables and the output power of the multiple fuel cell stacks as output variables to construct a fuzzy logic controller; Step S3: Establishing a hydrogen consumption function of the energy system based on the relationship between the output power and hydrogen consumption of the multiple fuel cell stacks and lithium battery stacks, and establishing an optimization function with the goal of minimizing the value of the hydrogen consumption function. The expression of the hydrogen consumption function is: H FCT (X)=H FC (X)+H bat (X); X=[x1 x2 x3 x4 x5 x6] T ; In the above formula, H FCT (X) is the hydrogen consumption function; H FC (X), H bat (X) are the equivalent hydrogen consumption functions of fuel cells and lithium batteries respectively; X is the decision vector X; x1 is the vertex, x2 is the lower left corner, x3 is the lower right corner, x4 is the lower scale, x5 is the lower left lag, and x6 is the lower right lag; α and β are weight matrices; α1, α2, α3, α4, α5, α6, β1, β2, β3, β4, β5, β6 are weight coefficients; H FC 、H bat Represent the hydrogen consumption of multiple fuel cell stacks and lithium battery respectively; P FC is the output power of multiple fuel cell stacks; η FC is the efficiency of multiple fuel cell stacks; P bat is the output power of the lithium battery; η bat The efficiency of lithium batteries; η c ,η d are the charging efficiency and discharging efficiency of lithium batteries respectively; t ini and t end Represent the initial time and the end time respectively; Step S4: Solving the optimization function to obtain optimal parameters of the fuzzy logic controller, and controlling the output power of the multiple fuel cell stacks according to the optimal parameters.

2. The energy management method for multiple fuel cell stacks based on fuzzy control according to claim 1, characterized in that: The construction of the fuzzy logic controller in step S2 includes the following steps: Step S21: defining a plurality of membership functions, and deriving the fuzzy sets of the input variables and the output variables according to the plurality of membership functions; Step S22: defining a mapping relationship between fuzzy sets of input variables and output variables as a fuzzy rule; Step S23: defining a plurality of free variables of the membership function, and using the free variables as parameters of the fuzzy logic controller.

3. The energy management method for multiple fuel cell stacks based on fuzzy control according to claim 2, characterized in that: In step S21 , seven membership functions mf1 - mf7 are constructed, where mf1 - mf7 correspond to very low, medium-low, low, medium, relatively high, medium-high, and very high, respectively.

4. The energy management method for multiple fuel cell stacks based on fuzzy control according to claim 2, characterized in that: The fuzzy rules in step S22 include: (1) When the state of charge of the lithium battery is low and the power demand of the energy system is also low, the fuel cell should output higher power; (2) When the state of charge of the lithium battery is low and the power demand of the energy system is high, the fuel cell should output higher power; (3) When the state of charge of the lithium battery is high and the power demand of the energy system is low, the fuel cell should output lower power; (4) When the state of charge of the lithium battery is high and the power demand of the energy system is also high, the fuel cell should output lower power.

5. The energy management method for multiple fuel cell stacks based on fuzzy control according to claim 2, characterized in that: The free variables in step S23 include vertex x1, lower left corner x2, lower right corner x3, lower scale x4, lower left lag x5 and lower right lag x6.

6. The energy management method for multiple fuel cell stacks based on fuzzy control according to claim 1, characterized in that: The expression for the output power of the multiple fuel cell stacks in step S3 is: P FC =V FC I FC =nV cell I FC ; In FC =nV cell ; V cell =E Nerest -V act -V ohm -V con ; V ohm =R FC I FC ; In the above formula, P FC is the output power of multiple fuel cell stacks; V FC is the output voltage of multiple fuel cell stacks; I FC is the output current of the multi-stack fuel cell; n is the number of fuel cells in the multi-stack fuel cell; V cell 、E Nerest 、V act 、V ohm 、V con 、T FC 、P H2 、P O2 、C O2 、R FC 、k con and I FC max are the output voltage, Nester voltage, activation polarization voltage drop, ohmic voltage drop, concentration polarization voltage drop, operating temperature, hydrogen pressure, oxygen pressure, oxygen concentration, internal resistance, concentration polarization voltage coefficient and maximum current of a single fuel cell respectively; ε1, ε2, ε3 and ε4 are constants respectively.

7. The energy management method for multiple fuel cell stacks based on fuzzy control according to claim 1, characterized in that: In step S4, the optimization function is solved by using a proxy model optimization algorithm and an adaptive particle swarm optimization algorithm, which includes the following steps: Step S41: Initializing the parameter set of the fuzzy logic controller and the parameters of the agent model optimization algorithm; Step S42: Calculate the hydrogen consumption corresponding to each set of parameters in the parameter set, and take the set of parameters with the minimum hydrogen consumption as the first optimal solution; Step S43: constructing a proxy model according to the current parameter set, interpolating the optimization function using the proxy model to obtain a proxy function, randomly generating a set of candidate solutions for the proxy function, and selecting a second optimal solution from the set of candidate solutions; Step S44: comparing the hydrogen consumption corresponding to the first optimal solution and the second optimal solution, taking the solution with the smaller hydrogen consumption as the current optimal solution, and adding the current optimal solution to the parameter set; Step S45: Repeat steps S42 to S44 until the current number of iterations reaches the set maximum number of iterations, and output the parameter set; Step S46: inputting the parameter set into an adaptive particle swarm optimization algorithm to obtain an optimal parameter combination.

8. The energy management method for multiple fuel cell stacks based on fuzzy control according to claim 7, characterized in that: The step S43 of selecting the second optimal solution from the candidate solution set includes the following steps: Step S431: Calculate each candidate point in the candidate solution set and sampling point set The shortest distance between Where, Represents the location of the points in the training set; Step S432: Set the maximum and minimum values ​​of the proxy function value S max 、S min , the maximum and minimum values ​​of the shortest distance between the candidate point and the sampling point D max 、D min for: Where, is the candidate solution set; Step S433: Generate an ascending weight set {σ1,σ2,…,σ κ }, if mod(τ,κ)≠0, then let σ S ←σ mod(τ,κ) , otherwise let σ S ←σ κ , where mod() is the modulo function; Step S434: Let σ D ←1-σ S ; Step S435: If S min ≠S max ,make Otherwise, Step S436: If D min ≠D max ,make Otherwise, Step S437: Update the current optimal solution: Step S438: Repeat steps S435 to S437 until the current number of iterations reaches the set maximum number of iterations, and output the current optimal solution as the second optimal solution.

9. An energy management system for multiple fuel cell stacks based on fuzzy control, implemented based on the energy management method for multiple fuel cell stacks based on fuzzy control according to any one of claims 1 to 8, characterized in that: include: Data acquisition module, energy management module and control module; The data acquisition module is used to collect operating data of multiple fuel cell stacks and lithium batteries; The energy management module allocates the output power of multiple fuel cells and lithium batteries according to the operating data collected by the data acquisition module to meet the power requirements of the system; The control module controls the output power of the multiple fuel cell stacks and lithium batteries according to the control signal output by the energy management module.

Images