A method for calculating the internal oxygen excess ratio of a fuel cell

Abstract

The present invention discloses a method for calculating the internal oxygen excess ratio of a fuel cell, characterized in that the method comprises the following steps: (1) establishing a sixth-order oxygen excess ratio characteristic model based on a nine-order model of the fuel cell according to assumptions of gas, humidity and temperature; (2) constructing a model estimation part of a high-order sliding mode observer based on this model; (3) performing parameter fusion and order reduction according to model characteristics, parameter approximation assumptions and the inlet gas components of the fuel cell, so as to reduce the operation cost of the output injection distribution matrix; (4) reconstructing the output injection vector by dynamically allocating the cathode gas output injection term to obtain the internal state estimation of the fuel cell, and further calculating the real-time oxygen excess ratio. The present invention can not only greatly reduce the operation cost of the sliding mode observer, but also has good dynamic response and small steady-state error, ensuring the real-time performance and accuracy of the oxygen excess ratio estimation of the fuel cell.

Description

Technical Field

[0001] The present invention relates to the technical field of fuel cells, and particularly to a method for calculating the oxygen excess ratio inside a fuel cell. Technical Background

[0002] The demand for reducing environmental pollution is increasing day by day, which is also driving the innovation of clean energy. Among them, the Proton Exchange Membrane Fuel Cell (PEMFC) is a typical representative of these clean energies and is considered to be one of the most promising technologies due to its potential high efficiency, high power density and high reliability.

[0003] However, for the PEMFC power generation system, since the cathode air supply speed lags far behind the electrochemical reaction speed when the load changes suddenly, this will cause the imbalance of the oxygen excess ratio (OER) inside the fuel cell. The oxygen excess ratio is defined as the ratio of the oxygen flow rate input to the fuel cell to the oxygen flow rate used for fuel cell power generation. When the OER is too low, the water generated at the cathode cannot be taken away in time, which affects the mass transfer of the gas in the membrane electrode at the diffusion layer, and even causes local hot spots to burn the membrane electrode, thus affecting the durability and reliability of the system. However, too high OER means an increase in the energy consumption of the air compressor, thus consuming more parasitic power and reducing the economy of the system. Therefore, it is necessary to control the OER in real time to avoid oxygen starvation and energy waste.

[0004] Before controlling the oxygen excess ratio, it is necessary to obtain the real-time oxygen excess ratio of the system. The acquisition of the oxygen excess ratio mainly uses methods such as flow sensors and observers. The sensors for measuring the air flow rate have a large delay, about 1 - 2 s. For the closed-loop control of the air system, the delay is relatively serious, which is bound to cause instability of the closed-loop control. In addition, the accuracy of the air flow sensor is about 1% - 10%, which will cause a large steady-state error in the entire control system. The observer methods mainly include extended observers, Kalman filter observers and sliding mode observers. The extended observer is simple to implement and requires fewer parameter variables, but it also has problems such as poor dynamic performance. The Kalman filter observer has excellent observation performance and small errors, but it is sensitive to noise and system uncertainties and has poor robustness. The convergence rate of the sliding mode observer is independent of the model parameters and external disturbances. Selecting a suitable error correction algorithm can eliminate the steady-state error and accelerate the system response speed. However, the order of its observability matrix is the same as the order of the system model and depends on the calculation of the Lie derivative, which makes the calculation cost of this method relatively large. Therefore, it is necessary to propose a calculation method with low operation cost and capable of calculating the oxygen excess ratio inside the fuel cell in real time. Summary of the Invention

[0005] In view of the deficiencies of the above-mentioned background technology, the present invention proposes a method for calculating the internal oxygen excess ratio of a fuel cell to achieve real-time calculation of the internal oxygen excess ratio of the fuel cell.

[0006] To achieve the above object, a method for estimating the internal oxygen amount of a fuel cell designed by the present invention includes the following steps:

[0007] S1: Establish a six-order model of the fuel cell oxygen excess ratio characteristic model, and the state variables of the oxygen excess ratio characteristic model are:

[0008]

[0009] x is the state vector in the state equation;

[0010] x1 represents the motor speed ω cp and x2 represents the intake pipe pressure P sm and x3 represents the intake pipe air quality m sm and x4 represents the oxygen mass inside the fuel cell stack x5 represents the nitrogen mass inside the fuel cell stack x6 represents the exhaust pipe pressure P rm ; It is easy to obtain the accurate values of x1, x2, and x6 using speed and pressure sensors; T is the matrix transpose;

[0011] S2: Construct a fuel cell high-order sliding mode observer model, which includes two parts: the six-order model in step S1 and the output distribution matrix. The output distribution matrix includes an observable matrix and a nonlinear error correction vector;

[0012] S3: Reduce the six-order observable matrix used in the output distribution matrix Lie derivative calculation in step S2 to a fifth order to obtain a reduced-order observable matrix;

[0013] S4: Reconstruct the output injection vector of the output distribution matrix in step S2 and calculate the internal oxygen excess ratio of the fuel cell.

[0014] Furthermore, the fuel cell high-order sliding mode observer model in step S2 is:

[0015]

[0016] Where: is the derivative of the state vector;

[0017] represents to is

[0018] a nonlinear equation;

[0019] ηcm - Mechanical efficiency;

[0020] K t - Motor constant 1, take 0.0225, unit: N·m·A -1 ;

[0021] J cp - Motor inertia, take 5×10 -5 , unit: Kg·m 2 ;

[0022] R cm - Motor constant 3, take 1.2, unit: ohm;

[0023] V cm - Air compressor motor voltage;

[0024] n - Number of single battery blocks, data value take 381;

[0025] - Molar mass of oxygen, take 28×10 -3 , unit Kg·mol -1 ;

[0026] F - Faraday constant;

[0027] I st - Fuel cell current;

[0028] θ(x) represents the output injection vector reconstruction of the fifth-order observable matrix.

[0029] represents the observable matrix;

[0030] v is the non-linear error correction vector.

[0031] Further, the calculation expression of the output distribution matrix in step S2 is:

[0032]

[0033] Where represents the observable matrix, and v represents the non-linear error correction vector.

[0034] Further, the non-linear error correction vector v = [v1 v2] T , T is the matrix transpose, and the calculation expression of the non-linear error correction term v1 is:

[0035] v1 = -α1U sign(e2 - β1·e 2,M )

[0036] β1 ∈ [0; 1)

[0037] where both v1 and v2 are non - linear error correction terms;

[0038] α and β are proportionality coefficients; α1 = 1, u = 0.1, β1 = 0.5;

[0039] e 2,M is the value of e2 at that time;

[0040] P sm is the actual intake - pipe pressure, and adding “^” to P sm represents the intake - pipe pressure obtained by the observer;

[0041] The calculation expression of the non - linear error correction term v2 obtained by using the quasi - continuous third - order sliding - mode algorithm is:

[0042]

[0043] where α2 = 10; P rm is the actual exhaust - pipe pressure, and adding “^” to P rm represents the exhaust - pipe pressure obtained by the observer.

[0044] Furthermore, in the step S3, the sixth - order observable matrix is reduced to the fifth - order. The involved reduction method includes: fusing the state variables of the sixth - order peroxide - ratio characteristic model, namely the oxygen mass inside the stack and the nitrogen mass inside the stack Specifically, in the state equation the oxygen mass inside the stack and the nitrogen mass inside the stack are removed, and then the oxygen mass inside the stack the nitrogen mass inside the stack and the water - vapor mass C are added together as the new state variable m of the reduced - order system ca , that is

[0045] where C represents the water - vapor mass inside the fuel cell under the given fuel - cell temperature and nitrogen gas constant; P v,ca is the saturated water - vapor pressure, V ca is the cathode volume of the stack, is the nitrogen constant, and T st is the fuel - cell temperature.

[0046] The expression form of the state variable of the obtained reduced - order model is:

[0047] X = [ω cp , P sm , msm , P rm , m ca T

[0048] where ω cp is the motor speed, P sm is the intake pipe pressure, m sm is the mass of air in the intake pipe, is the mass of oxygen and nitrogen inside the fuel cell stack; P rm is the exhaust pipe pressure, m ca is the mass of the cathode gas; T is the matrix transpose; P v,ca is the saturated vapor pressure, V ca is the volume of the cathode of the fuel cell stack, is the nitrogen constant, T st is the temperature of the fuel cell stack.

[0049] Furthermore, in the step S4, the calculation method of the oxygen excess ratio inside the fuel cell is as follows:

[0050]

[0051] where: is the real-time oxygen excess ratio of the system; respectively represent the state variables observed by the observer in step S2: the intake pipe pressure P sm , the mass of oxygen inside the fuel cell stack the mass of nitrogen inside the fuel cell stack K sm,out - the gas constant of the cathode inlet orifice;

[0052] n - the number of single cell blocks;

[0053] - the molar mass of oxygen;

[0054] F - the Faraday constant;

[0055] I st - the current of the fuel cell;

[0056] P v,ca - the saturated vapor pressure;

[0057] - the oxygen constant;

[0058] T st - the temperature of the fuel cell stack;

[0059] V ca - the volume of the cathode of the fuel cell stack;

[0060] - the molar mass of nitrogen;​

[0061] - Nitrogen constant.

[0062] The present invention also provides a device, including at least one processor; and a memory communicatively connected to the at least one processor; wherein, the memory stores instructions executable by the at least one processor, and when the instructions are executed by the at least one processor, the at least one processor is enabled to execute the above method.

[0063] The present invention further provides a computer-readable storage medium storing a computer program, characterized in that when the computer program is executed by a processor, the above method is implemented.

[0064] The beneficial effects of the present invention are as follows:

[0065] By reducing the sixth-order observation matrix to a fifth-order observation matrix, the present invention greatly reduces the operation cost of the observer, and the present case has better dynamic response and smaller steady-state error, ensuring the real-time performance and accuracy of the fuel cell over-oxygen ratio estimation. Description of the Drawings

[0066] Figure 1 It is a flowchart of the present invention. Detailed Embodiments

[0067] The following further describes the present invention in detail with reference to the drawings and embodiments, but the embodiments should not be construed as limiting the present invention.

[0068] The present invention is implemented based on the MATLAB / Simulink simulation platform. One simulation cycle is 30 s, the sampling time is 0.1 s, and the solution method used is the Runge-Kutta method. In this embodiment, the change trend of the fuel cell current is as follows: the initial current is 100 A, it becomes 180 A at 4 s, becomes 220 A at the 8th s, becomes 210 A at the 12th s, becomes 270 A at the 16th s, becomes 290 A at the 22nd s, and becomes 300 A at the 26th s.

[0069] Figure 1 The method for observing the internal state variables of a fuel cell constructed in the preferred embodiment of the present invention includes the following steps:

[0070] Step 1, based on the existing nine-order classical control model of a fuel cell, establish a sixth-order fuel cell over-oxygen ratio characteristic model:

[0071] The establishment of the over-oxygen ratio characteristic model is based on the following conditions:

[0072] (1) All gases such as air, oxygen, nitrogen, water vapor, and hydrogen are ideal gases;

[0073] (2) The internal gas component pressures, temperatures, and concentrations are evenly distributed;

[0074] (3) The temperature and humidity of the gas at the cathode inlet remain unchanged;

[0075] (4) The dynamic characteristics of the anode intake system are ignored;

[0076] (5) The internal temperature of the fuel cell is considered constant, and the heat generated by the reaction is discharged by the heat dissipation system.

[0077] The state variables of the peroxide ratio characteristic model are as follows:

[0078]

[0079] x is the state vector in the state equation;

[0080] x1 represents the rotational speed ω of the air compressor motor cp , x2 represents the intake pipe pressure P sm , x3 represents the air quality m in the intake pipe sm , x4 represents the oxygen mass inside the fuel cell stack x5 represents the nitrogen mass inside the fuel cell stack x6 represents the exhaust pipe pressure P rm ; The exact values of x1, x2, and x6 can be easily obtained using rotational speed and pressure sensors; T is the matrix transpose.

[0081] The state space equation of the peroxide ratio characteristic model can be expressed as:

[0082]

[0083] y = [h1, h2, h3] T = [x1, x2, x6] (2)

[0084] Where: is the derivative of the state vector;

[0085] f1(x) to f6(x) are non - linear equations;

[0086] V cm is the voltage of the air compressor motor;

[0087] I st is the current of the fuel cell;

[0088] The parameters represented by other characters are shown in Table 1,

[0089] y represents the output;

[0090] h1 is the conventional expression in the state equation. In the present invention, h1 is x1, h2 is x2, and h3 is x6; T is the matrix transpose, which transposes a row vector into a column vector.

[0091] The non-linear equations of f1(x) to f6(x) in formula (1) are as follows:

[0092]

[0093]

[0094] R in formula (2) and formula (6) a is the air constant and is a fixed value;

[0095] M in formula (6) and formula (7) a is the molar mass of air and is a fixed value;

[0096] R in formula (6) and formula (7) V is the gas constant of water vapor and is a fixed value;

[0097] P in formula (8) rm is the exhaust pipe pressure x6;

[0098] T in formula (8) rm is the temperature of the gas leaving the fuel cell stack;

[0099] V in formula (8) rm is the exhaust manifold volume and is a fixed value;

[0100] The parameters represented by other characters in formulas (3) to (8) are shown in Table 1.

[0101] The oxygen excess ratio is defined as the ratio of the oxygen flow rate input to the fuel cell to the oxygen flow rate used for fuel cell power generation. The oxygen excess ratio The specific expression is as follows:

[0102]

[0103] Formula (9) is an existing formula. From the expression of the oxygen excess ratio according to formula (9), it can be seen that the oxygen excess ratio is related to the masses of oxygen and nitrogen inside the fuel cell, but the masses of oxygen and nitrogen inside the fuel cell in the state vector are not easily obtained through sensors.

[0104] Some existing technologies obtain the masses of oxygen and nitrogen inside the fuel cell through a sliding mode observer. However, the existing observer has problems such as poor dynamic performance or high computational cost. The present invention designs a new type of high-order sliding mode observer model for fuel cells, which can obtain the masses of oxygen and nitrogen inside the fuel cell through the sliding mode observer, and then calculate the oxygen excess ratio through formula (9).

[0105] Table 1: Parameter Definition Table

[0106]

[0107]

[0108] Step 2: Based on the fuel cell oxygen excess ratio characteristic model in Step 1, construct a fuel cell high-order sliding mode observer model. The fuel cell high-order sliding mode observer includes two parts: the sixth-order model in S1 and the output distribution matrix. The output distribution matrix includes an observable matrix and an error correction vector;

[0109] Based on the fuel cell oxygen excess ratio characteristic model in Step 1, construct a fuel cell high-order sliding mode observer model. The established fuel cell high-order sliding mode observer model is a sixth-order model; the established fuel cell high-order sliding mode observer model is mainly for realizing the real-time observation of the oxygen excess ratio.

[0110] The state space equation of the established fuel cell high-order sliding mode observer model can be expressed as:

[0111]

[0112] Formula (10) adds a correction part on the basis of Formula (1). The correction part is the output distribution matrix Therefore, the established fuel cell high-order sliding mode observer includes two parts: the sixth-order model in Step 1 and the output distribution matrix.

[0113] In Formula (1) and Formula (10), the character with a "^" represents the value observed by the fuel cell observer, and the character without a "^" represents the actual value; in this embodiment, the values calculated by the observer are all represented by a "^".

[0114] y = [h1, h2, h3] T = [x1, x2, x6](11)

[0115] is the derivative of the state vector;

[0116] represents where T is the matrix transpose;

[0117] is the output distribution matrix, and the represents the observable matrix, and its matrix expression form is as follows:

[0118]

[0119] In the observable matrix, the Denotes the gradient operation on h1(x), and its specific form is as follows:

[0120] Denotes the r1-th Lie derivative of h1(x) with respect to f(x).

[0121] In the formula, r1, r2, and r3 are the observable indices of x1, x2, and x6 respectively, satisfying r1 + r2 + r3 = 6. Appropriate observable indices need to be selected to make nonsingular.

[0122] v is the nonlinear error correction vector, where v = [v1 v2] T , T is the matrix transpose, and the output estimation error used to calculate the nonlinear error correction vector is expressed as:

[0123]

[0124] Where adding a "^" on the character represents the value observed by the fuel cell observer, and the character without a "^" represents the actual value.

[0125] θ(x) represents the reconstruction of the output injection vector for the fifth-order observable matrix.

[0126] Step 3: Reduce the sixth-order observable matrix used for calculating the Lie derivative of the output distribution matrix in Step 2 to the fifth order to obtain the reduced-order observable matrix;

[0127] Since is composed of the Lie derivative rows, and its inverse matrix needs to be continuously calculated in subsequent iterative operations. On the premise of ensuring the existence of the inverse matrix, this operation process is complex and difficult to implement; in the oxygen ratio characteristic model of the sixth-order model fuel cell, the oxygen mass and nitrogen mass always appear in the form of . In the actual operation of the fuel cell, the air component flowing into the cathode side is mainly nitrogen, which also directly leads to the fact that generally the oxygen mass inside the fuel cell is less than the nitrogen mass. Numerically, oxygen oxygen constant is approximated to the nitrogen constant . Therefore, on the premise of ensuring the prediction accuracy, the following parameter fusion is made:

[0128]

[0129] In the original state space equation, x4 in f1(x)~f6(x) represents the oxygen mass inside the stack x5 represents the nitrogen mass inside the stack

[0130] The above parameter fusion is to use the state variable of the sixth-order oxygen ratio characteristic model, the oxygen mass inside the stack The mass of nitrogen inside the stack is removed, and then the mass of oxygen inside the stack The mass of nitrogen inside the stack and the mass of water vapor C are directly added together to form the new state variable m of the reduced-order system ca .

[0131] Then the state-space equation of the fuel cell can be reduced to fifth order, and its state variables are: X = [ω cp , P sm , m sm , P rm , m ca T = [x1, x2, x3, x6, x7] T , and the corresponding state-space equation is as follows:

[0132]

[0133] y = [h1, h2, h3] T = [x1, x2, x6](18)

[0134] where t i (x) (i = 1 - 5) represents a continuous vector function

[0135] Then the state-space equation of the fuel cell can be reduced to fifth order, and its state variables are: X' = [ω cp , P sm , m sm , P rm , m ca T = [x1, x2, x3, x6, x7] T The corresponding state-space equation is as follows:

[0136] x1 represents the motor speed ω cp ;

[0137] x2 represents the intake pipe pressure P sm ;

[0138] x3 represents the mass of air in the intake pipe m sm ;

[0139] x6 represents the exhaust pipe pressure P rm ;

[0140] x7 represents the new state vector m ca , m ca is the mass of oxygen inside the stack The mass of nitrogen inside the stack and the constant C representing the mass of water vapor are directly added together, that is, m​​ca

[0141]

[0142] Among them C represents the mass of water vapor inside the fuel cell at a given fuel cell temperature and nitrogen gas constant; where P v,ca is the saturated water vapor pressure, V ca is the volume of the cathode of the stack, is the nitrogen constant, T st is the temperature of the stack.

[0143] In the calculation of , according to the above fifth-order assumption, taking r1 = 0, r2 = 2, r3 = 3, then The specific expression of is as follows:

[0144]

[0145] Since the inverse of the observable matrix is required in the calculation, it is necessary to ensure its non-singularity. In actual operation and calculation, using the International System of Units will cause a large difference in the order of magnitude of the matrix elements, which may lead to the failure of the inverse matrix operation. Therefore, when calculating this matrix, we define the unit of the state variable elements as 100 rad / min, KPa, g, g, KPa respectively. Let's assume that x i represents the state variable in the International System of Units, and x i ' represents the state variable after unit correction, and its specific expression is as follows:

[0146]

[0147] [l1 l2 l3 l4 l5] = [10 -2 10 -3 10 3 10 3 10 -3 (21) The nonlinear error correction vector corresponding to the observer is:

[0148] v = [v1 v2] T (22)

[0149] v is the nonlinear error correction vector; both v1 and v2 are nonlinear error correction terms;

[0150] For r2 = 2, the generalized sub-optimal sliding mode algorithm with the sliding mode hyperplane of is used to obtain the nonlinear error correction term v1:

[0151]

[0152] where α and β are proportionality coefficients; α1 = 1, U = 0.1, β1 = 0.5, e x is an intermediate parameter;

[0153] e 2,M is the value of e2 at that time;

[0154] For r2 = 3, the sliding mode hyperplane is and the error correction term v2 is obtained by the quasi - continuous third - order sliding mode algorithm:

[0155]

[0156] where, P rm is the actual exhaust pipe pressure, and adding a "^" to P rm represents the exhaust pipe pressure obtained by observing through the observer.

[0157] α2 = 10, and the derivatives of each order of the error used in the formula are estimated by z 3,1 and z 3,2 respectively:

[0158]

[0159] Here, λ 2,0 = 15, λ 2,1 = 50, λ 2,2 = 10

[0160] Let be an intermediate parameter, and its specific expression is as follows:

[0161]

[0162] where represents the observable matrix, and its Lie derivative calculation model is the lumped parameter model of the fuel cell system after parameter fusion and order reduction.

[0163] The observable matrix is reduced to a 5×5 square matrix, reducing the difficulty of inverse matrix operation; however, the high - order sliding mode observer established in step two is of the sixth order, and the reduced - order observable matrix cannot directly participate in the output injection allocation operation. Therefore, output injection vector reconstruction is carried out; the output injection vector means that the observer is based on model correction, and the corrected part is defined as the output injection term (injection part) in the high - order sliding mode observer, that is, the output injection vector.

[0164] Assume that the output injection vector obtained from the fifth - order equation is: represents the reduced - order equation after The obtained output injection vector, the actual output injection vector is: It can be seen that unknown, which needs to be estimated through known vectors The value of, where represents the output injection vector actually required by the observer.

[0165] Step 4: θ(x) represents the reconstruction of the output injection vector for the fifth-order observable matrix in Step 3, and calculates the internal peroxide ratio of the fuel cell.

[0166] The reconstruction of the output injection vector specifically refers to using the super-twisting sliding mode algorithm to allocate the cathode gas mass error injection term, so as to obtain the output injection terms corresponding to the oxygen mass and nitrogen mass. In the sixth-order model of the fuel cell, the air compressor speed is also a state variable that can be easily observed by sensors. Therefore, the super-twisting sliding mode algorithm with is used to allocate the cathode gas mass output injection term, so as to obtain the output injection terms corresponding to the oxygen mass and nitrogen mass. The specific sixth-order output injection terms and through obtained are as follows:

[0167]

[0168] where the proportionality coefficient is α3 = 0.01, λ3 = 1

[0169] In order to prevent the system from becoming unstable due to excessive oxygen and nitrogen error injection terms in the dynamic response, the amplitude of the output injection term is restricted as follows:

[0170]

[0171] The constraint of the proportionality coefficient k can be obtained as follows:

[0172]

[0173] From this, the real-time peroxide ratio of the system can be calculated as:

[0174]

[0175] where: - The real-time peroxide ratio of the system;

[0176] represents the state variable observed through the observer;

[0177] P v,ca - The saturated pressure of water vapor, which is a fixed value;

[0178] I st - The fuel cell current;

[0179] The parameters represented by other characters are shown in Table 1.

[0180] The standard value of the internal oxygen excess ratio of a fuel cell should generally be around 2. If the error between the calculated real-time oxygen excess ratio and the standard value exceeds the preset range, the oxygen excess ratio can be adjusted by controlling the gas flow rate.

[0181] Based on the above method, the present invention further provides a device, including: at least one processor; and a memory communicatively connected to the at least one processor; wherein, the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the above method.

[0182] The present invention further provides a computer-readable storage medium storing a computer program, characterized in that the computer program, when executed by a processor, implements the above method.

[0183] In addition to the above embodiments, the present invention may have other embodiments. Any changes, modifications, substitutions, combinations, and simplifications made under any departure from the spirit and principle of the present invention shall be equivalent replacement methods and are all included in the protection scope required by the present invention.

Claims

1. A method for calculating the internal oxygen excess ratio of a fuel cell, characterized in that: The method includes the following steps: S1: Establish a six-order model of the fuel cell oxygen ratio characteristic model, and the state variables of the oxygen ratio characteristic model are: x is the state vector in the state equation; x1 represents the motor speed ωcp, x2 represents the intake pipe pressure Psm, x3 represents the intake pipe air mass msm, x4 represents the oxygen mass in the fuel cell stack x5 represents the nitrogen mass in the fuel cell stack x6 represents the exhaust pipe pressure Prm; T represents the matrix transpose; The state space equation of the oxygen ratio characteristic model is: y = [h1, h2, h3] T = [x1, x2, x6] Wherein: is the derivative of the state vector; f1(x) to f6(x) are non-linear equations; η cm is the mechanical efficiency; k t is the motor constant 1; J cp is the motor inertia; R cm is the motor constant 3; Vcm is the voltage of the air compressor motor; n is the number of single battery blocks; M O2 is the molar mass of oxygen; F is the Faraday constant; Ist- is the fuel cell current; y represents the output; h1 is the conventional expression in the state equation, h1 is x1, h2 is x2, and h3 is x6; S2: Construct a high-order sliding mode observer model for the fuel cell. The high-order sliding mode observer of the fuel cell includes two parts: the six-order model in step S1 and the output distribution matrix. The output distribution matrix includes an observable matrix and a non-linear error correction vector; The high-order sliding mode observer model of the fuel cell is: Wherein: is to take the derivative of the state vector; represent to is a non - linear equation; η cm is the mechanical efficiency; K t is the motor constant 1; J cp is the motor inertia; R cm is the motor constant 3; V cm is the air compressor motor voltage; n is the number of single - cell battery blocks; is the molar mass of oxygen; F is the Faraday constant; I st is the fuel cell current; θ(x) represents the output injection vector reconstruction of the fifth - order observable matrix; represents the observable matrix; v is the non - linear error correction vector; S3: Reduce the six-order observable matrix used in the Lie derivative calculation of the output distribution matrix in step S2 to a five-order to obtain a reduced-order observable matrix; The order reduction method involved in reducing the sixth-order observable matrix to the fifth order includes: regarding the state variables included in the sixth-order peroxide ratio characteristic model, the oxygen mass inside the stack and the nitrogen mass inside the stack for parameter fusion. Specifically, in the state equation the oxygen mass inside the stack and the nitrogen mass inside the stack are removed. Then, the oxygen mass inside the stack the nitrogen mass inside the stack and the water vapor mass C are directly added together as the new state variable m of the reduced-order system ca ; That is Among them C represents the mass of water vapor inside the fuel cell at a given fuel cell temperature and nitrogen gas constant; P v,ca is the saturated water vapor pressure, V ca is the volume of the cathode of the stack, is the nitrogen constant, T st is the temperature of the stack; The expression form of the state variables of the obtained reduced-order model is: X' = [ω cp , P sm , m sm , P rm , m ca T ​ where ω cp is the motor speed, P sm is the intake pipe pressure, m sm is the intake pipe air quality, is the oxygen mass inside the fuel cell stack, is the nitrogen mass inside the fuel cell stack, P rm is the exhaust pipe pressure, m ca is the cathode gas mass; T is the matrix transpose; P v,ca is the water vapor saturation pressure, V ca is the cathode volume of the fuel cell stack, is the nitrogen constant, T st is the fuel cell stack temperature; S4: Reconstruct the output injection vector for the output distribution matrix in step S2 and calculate the internal peroxide ratio of the fuel cell Wherein: System real-time peroxygen ratio; respectively represent the state variables observed by the observer in step S2 - intake pipe pressure P sm and the mass of oxygen inside the stack the mass of nitrogen inside the stack K sm,out - cathode inlet orifice gas constant; n - number of single cell blocks; - molar mass of oxygen; F - Faraday constant; I st - fuel cell current; P v,ca - water vapor saturation pressure; - oxygen constant; T st - stack temperature; V ca - stack cathode volume; - molar mass of nitrogen; - nitrogen constant.

2. The method for calculating the internal oxygen excess ratio of a fuel cell according to claim 1, characterized in that: The output allocation matrix expression in the step S2 is as follows: where represents the observable matrix, and v represents the non-linear error correction vector.

3. The method for calculating the internal peroxide ratio of a fuel cell according to claim 2, wherein: The non-linear error correction vector v = [v1 v2] T , where T represents matrix transpose. The calculation expression of the non-linear error correction term v1 is as follows: v1 = -α1U sign(e2 - β1·e 2,M ) β1∈[0；1) where v is the non-linear error correction vector; both v1 and v2 are non-linear error correction terms; T is the matrix transpose; α and β are proportionality coefficients; α1 = 1, U = 0.1, β1 = 0.5, e x is an intermediate parameter; e 2,M is the value of e2 at that time; P sm is the actual intake pipe pressure, P sm Adding "^" above indicates the intake pipe pressure obtained by observing through the observer; The calculation expression of the non-linear error correction term v2 is obtained by using the quasi-continuous third-order sliding mode algorithm: wherein α2 = 10; P rm is the actual exhaust duct pressure, P rm with a "^" added to represent the exhaust duct pressure obtained by observation through an observer.

4. A device, characterized in that, including: at least one processor; and a memory communicatively connected to the at least one processor; wherein, the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method according to any one of claims 1 to 3.

5. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, it implements the method according to any one of claims 1 to 3.

Images