CN117727977A - Hydrogen fuel cell air flow and humidity MIMO dynamic cooperative control method - Google Patents
Hydrogen fuel cell air flow and humidity MIMO dynamic cooperative control method Download PDFInfo
- Publication number
- CN117727977A CN117727977A CN202311718495.7A CN202311718495A CN117727977A CN 117727977 A CN117727977 A CN 117727977A CN 202311718495 A CN202311718495 A CN 202311718495A CN 117727977 A CN117727977 A CN 117727977A
- Authority
- CN
- China
- Prior art keywords
- humidity
- air flow
- fuel cell
- hydrogen fuel
- input
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 239000000446 fuel Substances 0.000 title claims abstract description 78
- UFHFLCQGNIYNRP-UHFFFAOYSA-N Hydrogen Chemical compound [H][H] UFHFLCQGNIYNRP-UHFFFAOYSA-N 0.000 title claims abstract description 68
- 239000001257 hydrogen Substances 0.000 title claims abstract description 68
- 229910052739 hydrogen Inorganic materials 0.000 title claims abstract description 68
- 238000000034 method Methods 0.000 title claims abstract description 61
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 claims abstract description 58
- 229910052760 oxygen Inorganic materials 0.000 claims abstract description 58
- 239000001301 oxygen Substances 0.000 claims abstract description 58
- 125000000864 peroxy group Chemical group O(O*)* 0.000 claims abstract description 19
- 230000008569 process Effects 0.000 claims abstract description 19
- 238000005312 nonlinear dynamic Methods 0.000 claims abstract description 13
- 210000004027 cell Anatomy 0.000 claims description 76
- 210000002569 neuron Anatomy 0.000 claims description 74
- 230000006870 function Effects 0.000 claims description 35
- 238000004422 calculation algorithm Methods 0.000 claims description 14
- 238000012937 correction Methods 0.000 claims description 12
- 230000003044 adaptive effect Effects 0.000 claims description 6
- 238000002474 experimental method Methods 0.000 claims description 6
- 150000002978 peroxides Chemical class 0.000 claims description 6
- 239000002131 composite material Substances 0.000 claims description 3
- 238000012545 processing Methods 0.000 claims description 3
- 230000000644 propagated effect Effects 0.000 claims 1
- 239000007789 gas Substances 0.000 abstract description 21
- 230000008859 change Effects 0.000 abstract description 10
- 230000004044 response Effects 0.000 abstract description 8
- 239000008358 core component Substances 0.000 abstract description 2
- 230000008278 dynamic mechanism Effects 0.000 abstract description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 36
- 229910001868 water Inorganic materials 0.000 description 22
- 238000006243 chemical reaction Methods 0.000 description 11
- 230000001276 controlling effect Effects 0.000 description 8
- 239000012528 membrane Substances 0.000 description 7
- 229920006395 saturated elastomer Polymers 0.000 description 7
- IJGRMHOSHXDMSA-UHFFFAOYSA-N Atomic nitrogen Chemical compound N#N IJGRMHOSHXDMSA-UHFFFAOYSA-N 0.000 description 4
- 230000004913 activation Effects 0.000 description 4
- 230000001133 acceleration Effects 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 230000007246 mechanism Effects 0.000 description 3
- 238000012546 transfer Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 239000008367 deionised water Substances 0.000 description 2
- 229910021641 deionized water Inorganic materials 0.000 description 2
- 230000003111 delayed effect Effects 0.000 description 2
- 230000004069 differentiation Effects 0.000 description 2
- 238000009792 diffusion process Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 239000012535 impurity Substances 0.000 description 2
- 230000010354 integration Effects 0.000 description 2
- 229910052757 nitrogen Inorganic materials 0.000 description 2
- 229910052698 phosphorus Inorganic materials 0.000 description 2
- 230000001105 regulatory effect Effects 0.000 description 2
- 230000035945 sensitivity Effects 0.000 description 2
- 238000004088 simulation Methods 0.000 description 2
- 229910052717 sulfur Inorganic materials 0.000 description 2
- 230000009897 systematic effect Effects 0.000 description 2
- 238000011217 control strategy Methods 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000003487 electrochemical reaction Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- JEGUKCSWCFPDGT-UHFFFAOYSA-N h2o hydrate Chemical compound O.O JEGUKCSWCFPDGT-UHFFFAOYSA-N 0.000 description 1
- 235000003642 hunger Nutrition 0.000 description 1
- 230000004043 responsiveness Effects 0.000 description 1
- 230000037351 starvation Effects 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E60/00—Enabling technologies; Technologies with a potential or indirect contribution to GHG emissions mitigation
- Y02E60/30—Hydrogen technology
- Y02E60/50—Fuel cells
Landscapes
- Fuel Cell (AREA)
Abstract
The invention discloses a dynamic cooperative control method for air flow and humidity MIMO of a hydrogen fuel cell, and belongs to the technical field of hydrogen fuel cell control. The complex structure and the dynamic mechanism of the oxygen supply system of the hydrogen fuel cell under the actual working condition are fully considered, and the dynamic processes of the core component air compressor, the humidifier and the cathode gas inlet and outlet are analyzed in detail. The method introduces a calibrated peroxy ratio reference set value and cathode relative humidity reference set value MAP table, and constructs a nonlinear dynamic model of the multi-input multi-output hydrogen fuel cell oxygen supply system with air flow and cathode humidity. The invention realizes the cooperation of the air compressor and the cathode humidity by adjusting and controlling the rotating speed of the air compressor and the power of the humidifier, and meets the index of quick and stable response of the load change of the hydrogen fuel cell. Compared with the prior art, the invention has the advantages of improving the efficiency of the hydrogen fuel cell and prolonging the service life.
Description
Technical Field
The invention relates to the technical field of hydrogen fuel cell control, in particular to a dynamic cooperative control method for air flow and humidity MIMO of a hydrogen fuel cell.
Background
With the continuous enhancement of environmental protection consciousness, new energy vehicles are attracting more and more attention as a clean and environment-friendly transportation means. The hydrogen fuel cell vehicle is used as a novel new energy vehicle, and has obvious advantages compared with the traditional vehicle. The hydrogen is used as an energy source, and the chemical energy is directly converted into electric energy through the fuel cell to drive the motor, so that the motor is not dependent on external charging, and the motor is started quickly. Meanwhile, the hydrogen fuel cell vehicle has no exhaust gas and noise emission in the running process, and is a real environment-friendly clean energy vehicle type. As a storable energy carrier, hydrogen energy is also advantageous for large-scale application of renewable energy. In a word, as a new generation new energy vehicle, the hydrogen fuel cell vehicle has obvious advantages in various aspects such as performance, environmental protection, practicability and the like, and has wide application prospect.
The hydrogen fuel cell vehicle is a vehicle driven by hydrogen fuel cell technology, and is characterized in that hydrogen and oxygen are subjected to electrochemical reaction to generate electric energy through a proton exchange membrane, so that a motor is driven to realize the operation of the vehicle. In a hydrogen fuel cell system, the control of air flow and humidity is critical to the performance and reliability of the system, and when the load is greatly changed, namely, the situation of sudden braking and sudden acceleration of a vehicle, if the oxygen supply system cannot timely adjust the supply of air flow and humidity, the phenomena of oxygen starvation, water flooding and the like of a cell stack can be caused, the efficiency of the fuel cell is reduced, and the service life of the cell stack is reduced. Therefore, by precisely controlling the air flow and the humidity, the performance consistency of the hydrogen fuel cell system can be realized, the stability and the reliability under different conditions are ensured, and the hydrogen fuel cell system is beneficial to being better applied to commercial battery vehicles and other scenes.
Patent document CN113644301a proposes a model-free adaptive sliding film and backstepping structure control method, which realizes control of air flow and pressure of a cathode of a galvanic pile by changing the rotation speed of a compressor and the opening degree of a back pressure valve. Patent document CN116314967a controls a humidifier in an oxygen supply system by a preset fuzzy control algorithm to realize humidity control of gas. However, the above control method only controls the flow rate or the cathode humidity singly, and has problems such as a delay in response speed, a complicated control algorithm, and difficulty in engineering and application, and the air flow rate and the cathode humidity have a strong coupling relationship, so that it is necessary to simultaneously consider the cooperative adjustment of the air flow rate and the cathode humidity in the hydrogen fuel cell system, and a dynamic cooperative control method for the air flow rate and the humidity MIMO of the hydrogen fuel cell is proposed.
Disclosure of Invention
The invention aims to solve the problems in the prior art and provides a dynamic cooperative control method for air flow and humidity MIMO of a hydrogen fuel cell, which realizes cooperative control on air flow and cathode humidity by adjusting and controlling the rotating speed of an air compressor and the power of a humidifier under the condition of keeping the cathode humidity of a cell stack of a hydrogen fuel cell system stable and meets the index of quick and stable response on the load change of the hydrogen fuel cell.
The technical scheme adopted by the invention for realizing the technical purpose is as follows: a dynamic cooperative control method for air flow and humidity MIMO of a hydrogen fuel cell comprises the following steps:
s1, constructing a multi-input multi-output nonlinear dynamic model of the oxygen supply system of the hydrogen fuel cell.
The fuel cell oxygen supply system comprises an air filter, an air compressor, an intercooler, a humidifier, a controller, a galvanic pile, a flowmeter and a humidity sensor, wherein external air sequentially passes through the air filter, the air compressor, the intercooler and the humidifier and finally enters the galvanic pile for reaction; the air filter is used for filtering impurities in input air, the air compressor is used for compressing the air to a proper pressure and controlling the flow of the input air, the intercooler is used for reducing the temperature of the compressed high-temperature air, the humidifier is used for humidifying the air to be fed into the electric pile, the flowmeter is used for detecting the air flow rate at the outlet of the electric pile, the humidity sensor is used for detecting the humidity of the electric pile, the air compressor controller and the humidity controller are respectively connected with the air compressor and the humidifier and used for regulating the rotating speed of the air compressor and the power of the humidifier, and the fuel cell oxygen supply system model comprises an air compressor dynamic model, a humidifier dynamic model and a cathode gas dynamic model;
the dynamic model of the air compressor can be obtained according to the mechanism of motor inertia and air compressor rotating speed:
wherein J is c Omega is the moment of inertia of the motor c Is the angular velocity of the air compressor, τ cm Is the input torque of the motor of the air compressor, τ c The load torque of the air compressor is;
input torque tau of air compressor motor is calculated by motor mechanism equation cm The method comprises the following steps:
wherein eta cm Representing the efficiency, k of the motor t And k v Sensitivity constant of input torque of air compressor and counter electromotive force constant of motor are respectively R cm For the resistance of the motor coil, v cm The input voltage of the air compressor;
the load torque of the air compressor is as follows by a thermodynamic formula:
wherein C is p Is the specific heat capacity of air, T atm 、P atm Atmospheric temperature and atmospheric pressure, η, respectively c Indicating the maximum efficiency of the air compressor, P sm Represents the outlet pressure of the air compressor, delta represents the specific heat capacity coefficient of air, Q c Air flow is the air flow of the outlet of the air compressor;
the dynamic characteristic equation of the angular speed of the air compressor can be arranged by the above formula:
the humidifier is used for humidifying the air before entering the cathode of the electric pile, and the air entering the humidifier can be divided into dry air Q dry,air,cl And water vapor Q H2O,cl ：
Q H2O,cl ＝Q cl -Q dry,air,cl
Wherein Q is cl Omega for air flow leaving the intercooler dry,air The calculation formula is as follows, for the percentage of dry air in the air:
wherein M is H2O And M air Representing the molar masses of water and air, P cl And P sat Saturated gas pressure and saturated steam pressure, RH, at the same temperature respectively leaving the intercooler cl Relative humidity of the air leaving the intercooler;
increased water vapor flow Q in humidifier H2O,in Is that
Wherein D is mem Represents the diffusion coefficient of water, θ is the thickness of the humidifier film, S w And S is a The water concentration of the deionized water side and the air side of the humidifier respectively, wherein A is the area of a humidifier film;
the thermodynamic change of the gas flow inside the humidifier is:
wherein,for the mass of water in the humidifier, +.>Is the constant pressure specific heat capacity of the water vapor, +.>P is the temperature of water in the humidifier hm For humidifier power, ">Indicating an increased water vapor flow in the humidifier, < >>Represents the enthalpy change of the water vapor, U represents the heat transfer coefficient, A α And A β Respectively representing the humidifier membrane area and the humidifier conduit area;
air flow rate Q of humidifier outlet hm ：
The dynamic model of the cathode gas of the galvanic pile is obtained by mass conservation, an ideal gas state equation and humidity characteristics, and oxygen Q in the cathode O2 Q, water H2O,ca The mass flow dynamic characteristic equation is:
wherein,and->Respectively, cathode input, cathode output and oxygen flow rate obtained by reaction and consumption in the galvanic pile>Respectively inputting and outputting water flow rate of the cathode, generating reaction and transmitting the water through the proton exchange membrane;
the dynamic equation of cathode humidity is:
wherein R is an ideal gas constant, T st To the temperature in the pile, V ca For the volume of the cathode, P sat,ca The cathode saturated vapor pressure at the same temperature.Is jointly determined by humidifier power and air compressor power, < >>Mainly determined by the power of the air compressor, < >>Is determined by the load current of the hydrogen fuel cell stack;
and constructing a multi-input multi-output nonlinear dynamic model of the fuel cell oxygen supply system based on the air compressor dynamic model, the humidifier dynamic model and the cathode gas dynamic model.
S2, calibrating a reference value MAP table of the peroxy ratio and a reference value MAP table of the cathode relative humidity with the maximum net power as a control target under different working conditions based on the multi-input multi-output nonlinear dynamic model of the oxygen supply system of the hydrogen fuel cell.
The peroxy ratio is generally the ratio of the flow of oxygen to the cathode of the stack to the flow of oxygen consumed by the stack reaction, and is a value greater than 1, calculated by the formula:
wherein,indicating the peroxy ratio>For the oxygen flow rate input to the cathode of the stack, < +.>Is the oxygen flow consumed by the galvanic pile reaction;
the oxygen flow rate consumed by the hydrogen fuel cell stack reaction can be calculated by the following formula:
wherein,is the molar mass of oxygen, N is the number of cells in the stack, I st Representing the load current of the galvanic pile, wherein F is Faraday constant;
air flow rate input to the cathode of the hydrogen fuel cell stack by the oxygen supply system:
wherein Q is air,ca,in For the air flow rate system input to the hydrogen fuel cell stack cathode, ω is the air humidity ratio,the oxygen mass fraction in the air;
wherein,is the mole fraction of oxygen, +.>Is the mole fraction of nitrogen;
net output power P of hydrogen fuel cell stack n The method comprises the following steps:
P n ＝P st -P c -P hm
wherein P is st For hydrogen fuel cell stack output, P c Power consumption for air compressor, P hm Consuming power for the humidifier;
s3, obtaining an air flow actual value and a cathode relative humidity actual value based on the MIMO nonlinear dynamic model of the hydrogen fuel cell oxygen supply system.
S4, referencing the air flow rate to a set value Q air,ca,in Air flow actual value Q air,ca,in Reference set point RH for cathode relative humidity * ca Actual value of cathode relative humidity RH ca Input to the adaptive dynamic network cooperative controller.
The self-adaptive dynamic network cooperative controller mainly comprises an input layer, a hidden layer and an output layer, and adopts error forward propagation and error backward propagation algorithms;
the error forward propagation algorithm forms the output of the network according to the input of the network, the weight coefficient, the state function and the output function of each layer of network, and the specific process is as follows:
the input layer is used for receiving feedback signals and target values from a system or a process, and comprises two sub-networks, each sub-network comprises two neurons, the input of the two neurons of one sub-network is respectively a reference set value of air flow and an actual value of air flow, the input of the two neurons of the other sub-network is respectively a reference set value of cathode relative humidity and an actual value of cathode relative humidity, and then the output of the input layer is as follows:
O li (1) (k)＝x li (k)
the hidden layer comprises two sub-networks for processing input signals and extracting features, each sub-network comprises a proportional neuron, an integral neuron and a differential neuron, and each neuron input is:
wherein net is lj (2) (k) As input of hidden layer neurons, w lij (2) As weighting coefficients of hidden layers, O li (1) (k) For output of the neurons of the input layer, l is the label of the sub-network, l=1, 2, i is the number of the neurons of the input layer of the sub-network, i=1, 2, j is the neurons of the hidden layer, j=1, 2,3 respectively represent the proportional neurons, the integral neurons and the differential neurons, and the upper right corner marks (1) (2) respectively represent the input layer and the hidden layer;
and S5, according to the self-adaptive dynamic network cooperative controller, acquiring proportional, differential and integral control parameter coefficients of the current moment Tbest, and further realizing dynamic cooperative control of the air compressor rotating speed and the humidifier power.
The proportional neuron adopts a unit proportional function as a state function u l1 (2) (k)：
u l1 (2) (k)＝net l1 (2) (k)
Wherein net is l1 (2) (k) The input value of the hidden layer proportion neuron at the moment k;
the state function u of the integral neuron l2 (2) (k) Expressed as a unit integral scale function:
u l2 (2) (k)＝u l2 (2) (k-1)+net l2 (2) (k)
wherein u is l2 (2) (k-1) is a state function value of the time-integral neuron of the hidden layer (k-1), net l2 (2) (k) Integrating the input value of the neuron for the hidden layer at the moment k;
the state function u of the differential neuron l3 (2) (k) Expressed as a unit differential proportional function:
u l3 (2) (k)＝net l3 (2) (k)-net l3 (2) (k-1)
wherein net is l3 (2) (k) For the input value of the hidden layer differential neuron at time k, net l3 (2) (k-1) is the input value of the hidden layer differential neuron at the time (k-1);
the hidden layer nonlinearly processes the input by the weight and activation function of the neural network, the hidden layer outputs O of each neuron lj (2) (k) The method comprises the following steps:
O lj (2) (k)＝f(u lj (2) (k))
f (x) is an activation function, and a Sigmoid function with positive and negative symmetry is taken;
the output layer is provided with two neurons, and the neurons of the output layer output net h (3) (k) The method comprises the following steps:
wherein w is ljh (3) For the weighting coefficients from hidden layer to output layer, h is the sequence number of the sub-network of the output layer, h=1, 2, o lj (2) (k) For the output of the hidden layer, the upper right corner mark (3) represents the output layer network;
the error back propagation algorithm adopts a gradient correction method to correct the weight, and continuously adjusts the network weight coefficient through network autonomous learning, so that the control quantity is infinitely close to the control target value, and the specific process of weight correction is as follows:
the weight coefficient w from the input layer to the hidden layer lij (2) (k) The iterative correction formula is:
e(k)＝y * (k)-y(k)
wherein eta is learning step length, J is deviation square mean value, e (k) is systematic error, y * (k) For prediction output, y (k) is a control target;
the weight coefficient w from the hidden layer to the output layer ljh (3) (k) The iterative correction formula is:
the self-adaptive dynamic network cooperative controller continuously adjusts the weight coefficient w through online autonomous learning lij (2) (k)、w ljh (3) (k) And the dynamic cooperative control of the controlled object is realized.
Compared with the prior art, the invention has the following beneficial effects:
1. the invention fully considers the complex structure and dynamic mechanism of the oxygen supply system of the hydrogen fuel cell under the actual working condition, and analyzes the dynamic process of the core component air compressor, the humidifier and the cathode gas in and out in detail; meanwhile, a calibrated peroxy ratio reference set value and a cathode relative humidity reference set value MAP table are introduced, and a nonlinear dynamic model of the multi-input multi-output hydrogen fuel cell oxygen supply system with air flow and cathode humidity is constructed;
2. aiming at the characteristics of a multi-input multi-output nonlinear dynamic model of a constructed hydrogen fuel cell oxygen supply system, the invention considers the engineering of practical application, and provides a multi-input multi-output self-adaptive cooperative control method combining proportion, differentiation and integration with a dynamic network, under the condition of keeping the cathode humidity of a stack of the hydrogen fuel cell system stable, the cooperative control of air flow and cathode humidity is realized by adjusting and controlling the rotating speed of an air compressor and the power of a humidifier, and the index of quick and stable response to the load change of the hydrogen fuel cell is satisfied;
3. the invention provides a dynamic cooperative control method for air flow and humidity MIMO of a hydrogen fuel cell, which aims to solve the problems that when load changes greatly, namely the conventional working conditions of sudden braking and sudden acceleration of a vehicle, the air flow or the cathode humidity is singly controlled, the response speed is delayed, the control algorithm is complex and engineering application is difficult, the adopted adaptive dynamic network cooperative control method can achieve a better generalized decoupling effect, network weight coefficients are automatically learned through a network, the actual control quantity is enabled to meet the asymptotically approximate set optimal target value, the tracking error accords with consistency, and the stable and rapid responsiveness and robustness of the oxygen supply system of the hydrogen fuel cell are effectively improved.
Drawings
FIG. 1 is a schematic diagram of the overall principle of the present invention;
FIG. 2 is a flow chart of an actual implementation of the present invention;
FIG. 3 is a block diagram of an algorithm for a hydrogen fuel cell adaptive dynamic network cooperative control method of the present invention;
fig. 4 is a physical structure diagram of the oxygen supply system of the hydrogen fuel cell of the present invention.
Detailed Description
The technical solutions of the embodiments of the present invention will be further fully and clearly described below with reference to the embodiments of the present invention and the accompanying drawings, and it is apparent that the described embodiments are only a part of, but not all embodiments of the present invention, and all embodiments obtained by those skilled in the art without making any inventive effort are within the scope of the present invention based on the embodiments of the present invention.
Example 1
As shown in fig. 1-4: a dynamic cooperative control method for air flow and humidity MIMO of a hydrogen fuel cell comprises the following steps:
calibrating a MAP (MAP) table of a peroxy ratio reference set value to obtain an air flow reference set value; calibrating a cathode relative humidity reference set value MAP table to obtain a cathode relative humidity reference set value; taking an air flow actual value, a cathode relative humidity actual value, an air flow reference set value and a cathode relative humidity reference set value in a multi-input multi-output nonlinear dynamic model of the fuel cell oxygen supply system as inputs of a composite dynamic network decoupling controller, so as to obtain the rotating speed of an air compressor and the power of a humidifier, and timely respond and adjust air flow and humidity supply, thereby realizing cooperative control of air flow and cathode humidity in the hydrogen fuel cell oxygen supply system;
the specific process for calibrating the MAP table of the peroxy ratio reference set value comprises the following steps:
the current of the pile is kept unchanged, the peroxy ratio is gradually changed from 1.5 to 3, and the net output power at each experimental point is calculated by using the current and voltage data of the pile. The stack current was then varied and the experiment repeated multiple times. The peroxide ratio corresponding to the maximum net output power is the reference setting value of the peroxide ratio under the current of the electric pileThe peroxy ratio reference set value, the corresponding pile current and the net output power are recorded in a MAP table to finish calibration, and the air flow reference set value Q required by an oxygen supply system can be obtained through the MAP table air,ca,in ；
The specific process for calibrating the MAP table of the reference set value of the relative humidity of the cathode comprises the following steps:
firstly, keeping cathode humidity unchanged, changing the magnitude of load current, recording the net output power of a galvanic pile under each experimental point, keeping the load current unchanged, changing the magnitude of cathode humidity, recording the net output power of the galvanic pile, repeating experiments for a plurality of times, wherein the cathode humidity corresponding to the maximum net output power is a cathode relative humidity reference set value under the galvanic pile current, and recording the cathode relative humidity reference set value, the corresponding galvanic pile current and the net output power in a MAP table to finish calibration;
the specific acquisition process of the air flow actual value and the cathode relative humidity actual value comprises the following steps:
and (3) constructing a hydrogen fuel cell oxygen supply system model, and respectively detecting an air flow actual value and a cathode relative humidity actual value through an air flow meter and a humidity sensor.
The self-adaptive dynamic network cooperative controller comprises: the input layer, the hidden layer and the output layer adopt error forward propagation and error backward propagation algorithms;
the error forward propagation forms the output of the network according to the input of the network, the weight coefficient of each layer of network, the state function and the output function.
A method for dynamically controlling the flow rate and humidity MIMO of a hydrogen fuel cell, referring to the flowcharts of fig. 1,2 and 3, comprising:
s1, constructing a multi-input multi-output nonlinear dynamic model of the oxygen supply system of the hydrogen fuel cell.
The fuel cell oxygen supply system is shown in fig. 4, and comprises an air filter, an air compressor, an intercooler, a humidifier, a controller, a galvanic pile, a flowmeter and a humidity sensor, wherein external air sequentially passes through the air filter, the air compressor, the intercooler and the humidifier and finally enters the galvanic pile for reaction, the air filter is used for filtering impurities in input air, the air compressor is used for compressing the air to a proper pressure and controlling the flow of the input air, the intercooler is used for reducing the temperature of the compressed high-temperature air, the humidifier is used for humidifying the air to be fed into the galvanic pile, the flowmeter is used for detecting the outlet air flow of the galvanic pile, the humidity sensor is used for detecting the humidity of the galvanic pile, the air compressor controller and the humidity controller are respectively connected with the air compressor and the humidifier and are used for regulating the rotating speed of the air compressor and the power of the humidifier, and the fuel cell oxygen supply system model comprises an air compressor dynamic model, a humidifier dynamic model and a cathode gas dynamic model;
specifically, the dynamic model of the air compressor can be obtained according to the relation between the inertia of the motor and the rotating speed of the air compressor, and the formula is as follows:
wherein J is c Omega is the moment of inertia of the motor c Is the angular velocity of the air compressor, τ cm Is the input torque of the motor of the air compressor, τ c The load torque of the air compressor is;
the input torque tau of the air compressor motor can be obtained by a motor mechanism equation cm The method comprises the following steps:
wherein eta cm Representing the efficiency, k of the motor t And k v Sensitivity constant of input torque of air compressor and counter electromotive force constant of motor are respectively R cm For the resistance of the motor coil, v cm The input voltage of the air compressor;
the thermodynamic formula is adopted, and the load torque of the air compressor is as follows:
wherein C is p Is the specific heat capacity of air, T atm 、P atm Atmospheric temperature and atmospheric pressure, η, respectively c Indicating the maximum efficiency of the air compressor, P sm Represents the outlet pressure of the air compressor, delta represents the specific heat capacity coefficient of air, Q c Air flow is the air flow of the outlet of the air compressor;
the dynamic characteristic equation of the angular speed of the air compressor can be arranged by the above formula:
the humidifier is used for humidifying the air before entering the cathode of the electric pile, and is used for reducing the temperature of the air before entering the cathode of the electric pileThe gas entering the humidifier may be divided into dry air Q dry,air,cl And steam
Wherein Q is cl Omega for air flow leaving the intercooler dry,air The calculation formula is as follows, for the percentage of dry air in the air:
wherein,and M air Representing the molar masses of water and air, P cl And P sat Saturated gas pressure and saturated steam pressure, RH, at the same temperature respectively leaving the intercooler cl Relative humidity of the air leaving the intercooler;
increasing water vapor flow in a humidifierIs that
Wherein D is mem Represents the diffusion coefficient of water, θ is the thickness of the humidifier film, S w And S is a Water concentrations on deionized water side and air side of humidifier, respectively, A α Is the area of the humidifier membrane;
the thermodynamic change of the gas flow inside the humidifier is:
wherein,for the mass of water in the humidifier, +.>Is the constant pressure specific heat capacity of the water vapor, +.>P is the temperature of water in the humidifier hm For humidifier power, ">Indicating an increased water vapor flow in the humidifier, < >>Represents the enthalpy change of the water vapor, U represents the heat transfer coefficient, A α And A β Respectively representing the humidifier membrane area and the humidifier conduit area;
humidifier outlet air flow Q hm ：
The dynamic model of the cathode gas of the galvanic pile is obtained by mass conservation, an ideal gas state equation and humidity characteristics, and oxygen in the cathodeWater->The mass flow dynamic characteristic equation of (2) is:
wherein,and->Respectively, cathode input, cathode output and oxygen flow rate obtained by reaction and consumption in the galvanic pile>Respectively inputting and outputting water flow rate of the cathode, generating reaction and transmitting the water through the proton exchange membrane;
specifically, the dynamic characteristic equation of cathode humidity is:
wherein R is an ideal gas constant, T st To the temperature in the pile, V ca For the volume of the cathode, P sat,ca Saturated vapor pressure at the same temperature as the cathode.Is jointly determined by humidifier power and air compressor power, < >>Mainly determined by the power of the air compressor, < >>Is determined by the load current of the galvanic pile;
and then building a multi-input multi-output nonlinear dynamic model of the fuel cell oxygen supply system based on the air compressor dynamic model, the humidifier dynamic model and the cathode gas dynamic model.
S2, calibrating a reference set value MAP table of the peroxy ratio and a reference set value MAP table of the cathode relative humidity with the maximum net power as a control target under different working conditions based on a multi-input multi-output nonlinear dynamic model of the oxygen supply system of the hydrogen fuel cell.
Firstly, calibrating a peroxy ratio reference set value MAP table, keeping the current of a pile unchanged, gradually changing the peroxy ratio from 1.5 to 3, and calculating the net output power at each experimental point by using current and voltage data of the pile. The stack current was then varied and the experiment repeated multiple times. The peroxide ratio corresponding to the maximum net output power is the reference setting value of the peroxide ratio under the current of the electric pileThe peroxy ratio reference set value, the corresponding pile current and the net output power are recorded in a MAP table to finish calibration, and the air flow reference set value Q required by an oxygen supply system can be obtained through the MAP table air,ca,in ；
The peroxy ratio is generally the ratio of the flow of oxygen to the cathode of the stack to the flow of oxygen consumed by the stack reaction, and is a value greater than 1, calculated by the formula:
wherein,indicating the peroxy ratio>For the oxygen flow rate input to the cathode of the stack, < +.>Is the oxygen flow consumed by the galvanic pile reaction;
the hydrogen fuel cell stack reactor consumed oxygen flow rate can be calculated as:
wherein,is the molar mass of oxygen, N is the number of cells in the stack, I st Representing the load current of the galvanic pile, wherein F is Faraday constant;
the oxygen supply system inputs the cathode air flow rate of the electric pile:
wherein Q is air,ca,in For the air flow rate input by the system to the stack cathode, ω is the air humidity ratio,the oxygen mass fraction in the air;
wherein,is the mole fraction of oxygen, +.>Is the mole fraction of nitrogen;
net output power P of hydrogen fuel cell stack n The method comprises the following steps:
P n ＝P st -P c -P hm
wherein P is st For the output power of the pile, P c Is the consumption power of the air compressor, P hm Power consumption for the humidifier;
specifically, calibrating a MAP table of a cathode relative humidity reference set value, firstly keeping the cathode humidity unchanged, changing the magnitude of load current, recording the net output power of the electric pile under each experimental point, then keeping the load current unchanged, changing the magnitude of the cathode humidity, recording the net output power of the electric pile, repeating the experiment for a plurality of times, wherein the cathode humidity corresponding to the maximum net output power is the RH of the cathode relative humidity reference set value under the current of the electric pile * ca Recording the cathode relative humidity reference set value, the corresponding pile current and the net output power in a MAP table, and completing calibration;
and (3) finishing a cathode humidity dynamic characteristic equation to obtain:
wherein R is an ideal gas constant, T st To the temperature in the pile, V ca For the volume of the cathode, P sat,ca Saturated vapor pressure at the same temperature as the cathode.Is jointly determined by humidifier power and air compressor power, < >>Mainly determined by the power of the air compressor, < >>Is determined by the load current of the galvanic pile;
the thermodynamic change of the gas flow inside the humidifier is:
wherein,for the mass of water in the humidifier, +.>Is the constant pressure specific heat capacity of the water vapor, +.>P is the temperature of water in the humidifier hm For humidifier power, ">Indicating increased water vapor flow in humidifier, H H2O Represents the enthalpy change of the water vapor, U represents the heat transfer coefficient, A α And A β Respectively representing the humidifier membrane area and the humidifier conduit area;
s3, obtaining an air flow actual value and a cathode relative humidity actual value based on the fuel cell oxygen supply system simulation model.
Specifically, the air flow actual value Q is obtained through a flowmeter in a simulation model of the fuel cell oxygen supply system air,ca,in The humidity sensor obtains the cathode relative humidity actual value RH ca ；
S4, referencing the air flow rate to a set value Q air,ca,in Air flow actual value Q air,ca,in Reference set point RH for cathode relative humidity * ca Actual value of cathode relative humidity RH ca Input to the adaptive dynamic network cooperative controller.
Specifically, the self-adaptive dynamic network cooperative controller mainly comprises an input layer, a hidden layer and an output layer, and adopts error forward propagation and error backward propagation algorithms;
the error forward propagation algorithm forms the output of the network according to the input of the network, the weight coefficient, the state function and the output function of each layer of network, and the specific process is as follows:
the input layer is used for receiving feedback signals and target values from the system or the process and comprises two sub-networks, each sub-network comprises two neurons, and the input of the two neurons of one sub-network is respectively the reference set value Q of the air flow air,ca,in And the actual value Q of the air flow air,ca,in The inputs of the two neurons of the other sub-network are respectively the reference set values RH of the cathode relative humidity * ca And the actual value RH of the relative humidity of the cathode ca Then the output of the input layer is:
O li (1) (k)＝x li (k)
the hidden layer comprises two sub-networks for processing input signals and extracting features, each sub-network comprises a proportional neuron, an integral neuron and a differential neuron, and each neuron input is:
wherein net is lj (2) (k) As input of hidden layer neurons, w lij (2) As weighting coefficients of hidden layers, O li (1) (k) For output of the neurons of the input layer, l is the label of the sub-network, l=1, 2, i is the number of the neurons of the input layer of the sub-network, i=1, 2, j is the neurons of the hidden layer, j=1, 2,3 respectively represent the proportional neurons, the integral neurons and the differential neurons, and the upper right corner marks (1) (2) respectively represent the input layer and the hidden layer;
and S5, acquiring a current proportional coefficient, a differential coefficient and an integral coefficient according to the composite neural network controller, and further controlling the air compressor rotating speed and the humidifier power.
The proportional neuron adopts a unit proportional function as a state function u l1 (2) (k)：
u l1 (2) (k)＝net l1 (2) (k)
Wherein net is l1 (2) (k) The input value of the hidden layer proportion neuron at the moment k;
the state function u of the integral neuron l2 (2) (k) Expressed as a unit integral scale function:
u l2 (2) (k)＝u l2 (2) (k-1)+net l2 (2) (k)
wherein u is l2 (2) (k-1) is a state function value of the time-integral neuron of the hidden layer (k-1), net l2 (2) (k) Integrating the input value of the neuron for the hidden layer at the moment k;
the state function u of the differential neuron l3 (2) (k) Expressed as a unit differential proportional function:
u l3 (2) (k)＝net l3 (2) (k)-net l3 (2) (k-1)
wherein net is l3 (2) (k) For the input value of the hidden layer differential neuron at time k, net l3 (2) (k-1) is the input value of the hidden layer differential neuron at the time (k-1);
the hidden layer nonlinearly processes the input by the weight and activation function of the neural network, the hidden layer outputs O of each neuron lj (2) (k) The method comprises the following steps:
O lj (2) (k)＝f(u lj (2) (k))
f (x) is an activation function, and a Sigmoid function with positive and negative symmetry is taken;
the output layer has two neurons, and the output net of the neurons of the output layer h (3) (k) The method comprises the following steps:
wherein w is ljh (3) For the weighting coefficients from hidden layer to output layer, h is the sequence number of the sub-network of the output layer, h=1, 2, o lj (2) (k) For the output of the hidden layer, the upper right corner mark (3) represents the output layer network;
the error back propagation algorithm adopts a gradient correction method to correct the weight, and continuously adjusts the network weight coefficient through network autonomous learning, so that the control quantity is infinitely close to the control target value, and the specific process of weight correction is as follows:
the weight coefficient w from the input layer to the hidden layer lij (2) (k) The iterative correction formula is:
e(k)＝y * (k)-y(k)
wherein eta is learning step length, J is deviation square mean value, e (k) is systematic error, y * (k) For prediction output, y (k) is a control target;
the weight coefficient w from the hidden layer to the output layer ljh (3) (k) The iterative correction formula is:
the self-adaptive dynamic network cooperative controller continuously adjusts the weight coefficient w through online autonomous learning lij (2) (k)、w ljh (3) (k) Realizing the control of the controlled object;
aiming at the characteristics of a constructed multi-input multi-output nonlinear dynamic model of the oxygen supply system of the hydrogen fuel cell, the invention considers the engineering of practical application, provides a multi-input multi-output cooperative control strategy combining proportion, differentiation and integration with a dynamic network, and realizes the cooperative control of the air compressor and the cathode humidity by adjusting and controlling the rotating speed and the humidifier power under the condition of keeping the cathode humidity of a galvanic pile of the hydrogen fuel cell system stable, thereby meeting the index of quick and stable response to the load change of the hydrogen fuel cell.
The invention aims to solve the problems that when the load changes greatly, namely the conventional working conditions of sudden braking and sudden acceleration of a vehicle, the air flow or cathode humidity is controlled singly, the response speed is delayed, the control algorithm is complex and difficult to engineer and apply in the oxygen supply system of the fuel cell, a self-adaptive dynamic network cooperative control method is adopted to achieve a better generalized decoupling effect, the network weight coefficient is continuously adjusted through network autonomous learning, the actual control quantity meets the asymptotically approaching set optimal target value, the tracking error accords with consistency, and the stable and rapid response and robustness of the oxygen supply system of the hydrogen fuel cell are effectively improved.
Claims (13)
1. A method for dynamically controlling the flow rate and humidity MIMO of a hydrogen fuel cell, comprising:
calibrating a MAP (MAP) table of a peroxy ratio reference set value to obtain an air flow reference set value;
calibrating a cathode relative humidity reference set value MAP table to obtain a cathode relative humidity reference set value;
the air flow actual value, the cathode relative humidity actual value, the air flow reference set value and the cathode relative humidity reference set value in a multi-input multi-output nonlinear dynamic model of the fuel cell oxygen supply system are used as the input of a composite dynamic network decoupling controller, so that the air compressor rotating speed and the humidifier power are obtained, the air flow and the humidity supply are responded and adjusted in time, and the cooperative control of the air flow and the cathode humidity in the hydrogen fuel cell oxygen supply system is realized.
2. The method for dynamically controlling the air flow rate and the humidity MIMO of the hydrogen fuel cell according to claim 1, wherein the specific process of calibrating the MAP table of the peroxy ratio reference set value is as follows:
keeping the current of the electric pile unchanged, gradually changing the peroxy ratio from 1.5 to 3, and calculating the net output power at each experimental point by using the current and voltage data of the electric pile; then changing the current of the electric pile, and repeating the experiment for a plurality of times; the peroxide ratio corresponding to the maximum net output power is the reference setting value of the peroxide ratio under the current of the electric pileThe peroxy ratio reference set value, the corresponding pile current and the net output power are recorded in a MAP table to finish calibration, and the air flow reference set value Q required by an oxygen supply system can be obtained through the MAP table air,ca,in 。
3. The method for dynamically controlling the air flow rate and the humidity MIMO of the hydrogen fuel cell according to claim 1, wherein the specific process of calibrating the MAP table of the reference set value of the cathode relative humidity is as follows:
firstly, keeping cathode humidity unchanged, changing the magnitude of load current, recording the net output power of a galvanic pile under each experimental point, keeping the load current unchanged, changing the magnitude of cathode humidity, recording the net output power of the galvanic pile, repeating the experiment for a plurality of times, wherein the cathode humidity corresponding to the maximum net output power is the cathode relative humidity reference set value under the galvanic pile current, and recording the cathode relative humidity reference set value, the corresponding galvanic pile current and the net output power in a MAP table to finish calibration.
4. The method for dynamically controlling the air flow and the humidity MIMO of the hydrogen fuel cell according to claim 1, wherein the specific process for obtaining the air flow actual value and the cathode relative humidity actual value is as follows:
and (3) constructing a hydrogen fuel cell oxygen supply system model, and respectively detecting an air flow actual value and a cathode relative humidity actual value through an air flow meter and a humidity sensor.
5. The method for dynamically cooperative control of air flow and humidity MIMO of a hydrogen fuel cell according to claim 1, wherein said adaptive dynamic network cooperative controller comprises: the input layer, the hiding layer and the output layer adopt error forward propagation and error backward propagation algorithms to complete input, hiding and output.
6. The method for dynamically coordinated control of air flow and humidity MIMO of a hydrogen fuel cell according to claim 5, wherein said error is propagated forward to form an output of the network based on the input of the network, the weight coefficients of the layers of the network, the status function and the output function.
7. The method for dynamically controlling the air flow and the humidity MIMO of the hydrogen fuel cell according to claim 5, wherein the error back propagation algorithm adopts a gradient correction method to correct the weight, the network weight coefficient is continuously adjusted through network autonomous learning, the control quantity is infinitely close to the control target value, and the specific process of the weight correction is as follows:
the weight coefficient w from the input layer to the hidden layer lij (2) (k) The iterative correction formula is:
wherein eta is learning step length, J is deviation square mean value, y * (k) For prediction output, y (k) is a control target;
the weight coefficient w from the hidden layer to the output layer ljh (3) (k) The iterative correction formula is:
the self-adaptive dynamic network cooperative controller continuously adjusts the weight coefficient w through online autonomous learning lij (2) (k)、w ljh (3) (k) And the control of the controlled object is realized.
8. The method for dynamically coordinated control of hydrogen fuel cell air flow and humidity MIMO of claim 5, wherein said input layer is configured to receive feedback signals and target values from a system or process, comprising: two sub-networks, each sub-network comprises two neurons, wherein the input of the two neurons of one sub-network is respectively a reference set value of air flow and an actual value of air flow, the input of the two neurons of the other sub-network is respectively a reference set value of cathode relative humidity and an actual value of cathode relative humidity, and the output of the input layer is as follows:
O li (1) (k)＝x li (k)。
9. the method of claim 5, wherein the hidden layer comprises two sub-networks for processing the input signal and extracting the features, each sub-network comprising a proportional neuron, an integral neuron and a derivative neuron, and the inputs of each neuron are:
wherein net is lj (2) (k) As input of hidden layer neurons, w lij (2) As weighting coefficients of hidden layers, O li (1) (k) For the output of the input layer neurons, l is the label of the sub-network, l=1, 2, i is the neuron number of the sub-network input layer, i=1, 2, j is the neuron of the hidden layer, j=1, 2,3 represents the proportional neuron, the integral neuron and the differential neuron, respectively, and the upper right corner marks (1) (2) represent the input layer and the hidden layer, respectively.
10. The method for dynamically controlling air flow and humidity MIMO of a hydrogen fuel cell according to claim 5, wherein said output layer comprises: two neurons, the output net of the output layer neuron h (3) (k) The method comprises the following steps:
wherein w is ljh (3) For the weighting coefficients from hidden layer to output layer, h is the sequence number of the sub-network of the output layer, h=1, 2, o lj (2) (k) For implicit layer output, the upper right corner mark (3) represents the output layer network.
11. The method for dynamically controlling air flow and humidity MIMO according to claim 9, wherein said proportional neurons use a unit proportional function as a state function u l1 (2) (k)：
u l1 (2) (k)＝net l1 (2) (k)
Wherein net is l1 (2) (k) And (5) implying the input value of the layer proportion neuron for the k moment.
12. The method for dynamically controlling air flow and humidity MIMO according to claim 9, wherein the state function u of the integral neuron l2 (2) (k) Expressed as a unit integral scale function:
u l2 (2) (k)＝u l2 (2) (k-1)+net l2 (2) (k)
wherein u is l2 (2) (k-1) is a state function value of the time-integral neuron of the hidden layer (k-1), net l2 (2) (k) The input values of the neurons are integrated for the hidden layer at time k.
13. The method for dynamically controlling air flow and humidity MIMO according to claim 9, wherein said differential neuron state function u l3 (2) (k) Expressed as a unit differential proportional function:
u l3 (2) (k)＝net l3 (2) (k)-net l3 (2) (k-1)
wherein net is l3 (2) (k) Input values for hidden layer differential neurons at time k，net l3 (2) (k-1) is the input value of the hidden layer differential neuron at the time (k-1).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311718495.7A CN117727977A (en) | 2023-12-14 | 2023-12-14 | Hydrogen fuel cell air flow and humidity MIMO dynamic cooperative control method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311718495.7A CN117727977A (en) | 2023-12-14 | 2023-12-14 | Hydrogen fuel cell air flow and humidity MIMO dynamic cooperative control method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117727977A true CN117727977A (en) | 2024-03-19 |
Family
ID=90206485
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311718495.7A Pending CN117727977A (en) | 2023-12-14 | 2023-12-14 | Hydrogen fuel cell air flow and humidity MIMO dynamic cooperative control method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117727977A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN119148770A (en) * | 2024-11-18 | 2024-12-17 | 上海扬基电子科技有限公司 | Intelligent control system and method for oil special-purpose explosion-proof digital display turbine flowmeter |
-
2023
- 2023-12-14 CN CN202311718495.7A patent/CN117727977A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN119148770A (en) * | 2024-11-18 | 2024-12-17 | 上海扬基电子科技有限公司 | Intelligent control system and method for oil special-purpose explosion-proof digital display turbine flowmeter |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112072142B (en) | Fuel cell control method and system based on model predictive control | |
| CN111342086B (en) | A fuel cell air-peroxygen ratio and flow pressure coordinated control method and system | |
| Wang et al. | Observer-based discrete adaptive neural network control for automotive PEMFC air-feed subsystem | |
| CN108134114A (en) | A kind of PEMFC Temperature control method | |
| CN108091909A (en) | It is a kind of based on optimal peroxide than fuel battery air flow control methods | |
| CN103560261A (en) | Control method of membrane humidifier for proton exchange membrane fuel cell | |
| CN110069033A (en) | A kind of total power fuel cell air compressor machine bilayer forecast Control Algorithm | |
| CN117727977A (en) | Hydrogen fuel cell air flow and humidity MIMO dynamic cooperative control method | |
| CN116435557A (en) | A neural network-based fuel cell thermal management method, device and system | |
| CN117518780B (en) | Fuel cell cogeneration system control method based on simulation model | |
| CN113442795B (en) | Control method of fuel cell hybrid power system based on layered MPC | |
| CN112713288B (en) | A fuel cell bubbling humidifier control system and control method | |
| CN112436166A (en) | Sliding mode control method of air supply system of proton exchange membrane fuel cell | |
| CN113644301B (en) | Cathode flow and pressure control method for vehicle fuel cell | |
| CN111769312B (en) | Fuel cell supply path decoupling control method based on pressure compensation | |
| Chen et al. | Improve hydrogen economy for vehicular fuel cell system via investigation and control of optimal operating oxygen excess ratio | |
| CN114976128A (en) | PEMFC fault prediction and health management system and method based on five-dimensional digital twin technology | |
| CN114488821B (en) | Method and system for predicting and controlling interval economic model of fuel cell oxygen passing ratio | |
| CN114857061B (en) | A Modeling and Multi-objective Control Method for Aviation Fuel Cell Air Supply System | |
| Song et al. | AI-based proton exchange membrane fuel cell inlet relative humidity control | |
| Xue et al. | An optimized Fuzzy PI control method utilizing an improved QPSO for the hydrogen supply of PEMFC | |
| Tao et al. | Nonlinear modeling of PEMFC based on neural networks identification | |
| CN114530618B (en) | Matching modeling method of fuel cell and air compressor based on stochastic optimization algorithm | |
| Li et al. | Observer-based control of PEMFC air feed systems via high-order fully actuated system approaches | |
| CN116565271A (en) | A control strategy for water-cooled fuel cell air supply system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |