WO2023216150A1 - 一种燃料电池的热管理方法 - Google Patents

一种燃料电池的热管理方法 Download PDF

Info

Publication number
WO2023216150A1
WO2023216150A1 PCT/CN2022/092245 CN2022092245W WO2023216150A1 WO 2023216150 A1 WO2023216150 A1 WO 2023216150A1 CN 2022092245 W CN2022092245 W CN 2022092245W WO 2023216150 A1 WO2023216150 A1 WO 2023216150A1
Authority
WO
WIPO (PCT)
Prior art keywords
fuzzy
fuzzy controller
particle
temperature
thermal management
Prior art date
Application number
PCT/CN2022/092245
Other languages
English (en)
French (fr)
Inventor
郑春花
伏圣祥
Original Assignee
中国科学院深圳先进技术研究院
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 中国科学院深圳先进技术研究院 filed Critical 中国科学院深圳先进技术研究院
Priority to PCT/CN2022/092245 priority Critical patent/WO2023216150A1/zh
Publication of WO2023216150A1 publication Critical patent/WO2023216150A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H01—ELECTRIC ELEMENTS
    • H01M—PROCESSES OR MEANS, e.g. BATTERIES, FOR THE DIRECT CONVERSION OF CHEMICAL ENERGY INTO ELECTRICAL ENERGY
    • H01M8/00—Fuel cells; Manufacture thereof
    • H01M8/04—Auxiliary arrangements, e.g. for control of pressure or for circulation of fluids
    • H01M8/04298—Processes for controlling fuel cells or fuel cell systems
    • H01M8/04694—Processes for controlling fuel cells or fuel cell systems characterised by variables to be controlled
    • H01M8/04858—Electric variables
    • H—ELECTRICITY
    • H01—ELECTRIC ELEMENTS
    • H01M—PROCESSES OR MEANS, e.g. BATTERIES, FOR THE DIRECT CONVERSION OF CHEMICAL ENERGY INTO ELECTRICAL ENERGY
    • H01M8/00—Fuel cells; Manufacture thereof
    • H01M8/04—Auxiliary arrangements, e.g. for control of pressure or for circulation of fluids
    • H01M8/04298—Processes for controlling fuel cells or fuel cell systems
    • H01M8/04992—Processes for controlling fuel cells or fuel cell systems characterised by the implementation of mathematical or computational algorithms, e.g. feedback control loops, fuzzy logic, neural networks or artificial intelligence

Definitions

  • the present invention relates to the technical field of fuel cell management, and more specifically, to a thermal management method for a fuel cell.
  • Hydrogen energy is a secondary energy source with rich sources, green, low-carbon and widely used. It is of great significance to building a clean, low-carbon, safe and efficient energy system and achieving the goal of peak carbon neutrality. With the rapid development of new energy vehicles, hydrogen fuel cell vehicles have received widespread attention due to their advantages such as high efficiency and cleanliness.
  • Proton Exchange Membrane Fuel Cell (PEMFC) has the advantages of high energy conversion efficiency, low temperature operation, high reliability and zero emissions, and has broad application prospects in the automotive field.
  • PEMFC is a nonlinear complex system with multi-physics and multi-parameter coupling. Its operating temperature is a key factor affecting output performance and life.
  • PEMFC thermal management can be divided into three methods: structure adjustment, phase change cooling and optimization control according to the principle. Changing the flow channel structure will complicate the internal structure of the fuel cell, increase the temperature control cycle cooling shell, and increase the volume of the fuel cell.
  • the fluid used in phase change cooling is relatively expensive and is not conducive to commercialization.
  • Both application methods have shortcomings.
  • PEMFC thermal management mainly controls the cooling water flow rate and fan speed based on the temperature model.
  • the control methods include PI (Proportion Integration) control, state feedback control, predictive control, etc. These control methods are simple in principle and easy to use, but have shortcomings such as slow response speed and long adjustment time.
  • the existing fuzzy control design mainly relies on the experience of experts, and most methods use step load signals to verify fuzzy control.
  • hydrogen fuel cell vehicles will have processes such as acceleration, uniform speed, and deceleration during actual driving.
  • Frequent changes in working conditions will make fuel cell temperature control more complex, and the temperature may have extreme values.
  • Existing fuzzy logic still has difficulties in controlling the temperature to the target value, and the control target temperature value fluctuates greatly.
  • the object of the present invention is to overcome the above-mentioned shortcomings of the prior art and provide a thermal management method for a fuel cell, which method includes the following steps:
  • an outlet fuzzy controller and an inlet fuzzy controller are set.
  • the fuel cell thermal management system includes a water tank, a cooling water pump, a radiator and a stack.
  • the outlet fuzzy controller uses the current stack outlet temperature and The error and error change rate of the set target outlet temperature are used as the input quantity, and the cooling water flow rate is used as the output quantity;
  • the inlet fuzzy controller uses the error and the error change rate between the current stack inlet temperature and the set target inlet temperature.
  • the radiator fan speed is used as the output quantity;
  • the membership function is parameterized, and the parameters to be optimized are used as the particle population to optimize the membership functions of the inlet fuzzy controller and the outlet fuzzy controller to control the current cooling water flow of the fuel cell thermal management system. and radiator fan speed.
  • the advantage of the present invention is that it aims at smaller errors between the inlet and outlet temperature of the stack and the target temperature value, and combines the characteristics of fuzzy control rules and membership function optimization to achieve better overall performance.
  • the algorithm optimizes the membership function of the fuzzy controller.
  • the optimized fuzzy controller has significantly improved stability and accuracy compared to the conventional fuzzy controller, overcoming the defect that the particle swarm optimization algorithm easily falls into local optimality.
  • the present invention has better temperature adjustment capabilities, smaller deviations from the set value, and can better resist disturbances from external loads.
  • Figure 1 is a schematic diagram of a fuel cell thermal management system in the prior art
  • Figure 2 is a flow chart of a thermal management method for a fuel cell according to an embodiment of the present invention
  • Figure 3 is a schematic diagram of the fuzzy control process according to an embodiment of the present invention.
  • Figure 4 is a schematic diagram of a triangular membership function according to an embodiment of the present invention.
  • Figure 5 is a schematic diagram of a fuzzy controller optimized based on a particle swarm-genetic hybrid algorithm according to an embodiment of the present invention
  • Figure 6 is a particle swarm-genetic hybrid algorithm optimized fuzzy control flow chart according to an embodiment of the present invention.
  • Figure 7 is a schematic diagram of the chromosome crossover process according to an embodiment of the present invention.
  • any specific values are to be construed as illustrative only and not as limiting. Accordingly, other examples of the exemplary embodiments may have different values.
  • the PEMFC thermal management system is first introduced, as shown in Figure 1.
  • the system includes a water tank, a cooling water pump, a radiator, and a stack (i.e., a fuel cell stack) with temperature sensors installed at the entrances and exits of the stack.
  • the generated heat is first brought to the water tank by the cooling water pump by controlling the cooling water flow, and then the heat is taken to the radiator.
  • the radiator discharges the heat into the air by controlling the radiator air volume.
  • the temperature of the cooling water is uniform, the temperature of the cooling water at the outlet of the stack is taken as the temperature at the outlet of the stack, and the outlet temperature of the radiator is taken as the temperature at the inlet of the stack.
  • the provided thermal management method of the fuel cell includes the following steps:
  • Step S110 For the fuel cell thermal management system, use the stack as a reference to set the outlet fuzzy controller and the inlet fuzzy controller.
  • two two-dimensional fuzzy controllers are used to control the temperature of the inlet and outlet of the stack. According to the selected stack, the outlet target temperature T tar.out is set, the stack inlet target temperature T tar.in is set, and the fuzzy inference system is designed using appropriate fuzzy control rules. After fuzzy inference, the weighted average method can be used for defuzzification.
  • Fuzzy control is essentially a nonlinear control and belongs to the category of intelligent control.
  • two mandatory-type two-dimensional fuzzy controllers called outlet fuzzy controller and inlet fuzzy controller respectively, are established to control the stack outlet temperature and inlet temperature.
  • the overall fuzzy control framework As shown in Figure 3.
  • For the stack outlet temperature control set the stack outlet target temperature T tar.out , and set the error and error change rate between the current stack outlet temperature T st.out and the target outlet temperature T tar.out as the output fuzzy controller.
  • the cooling water flow rate is used as the output of the outlet fuzzy controller.
  • the heat generated by the stack is first brought to the water tank through the cooling water to reach the target temperature value of the stack outlet; after a part of the heat is lost, the remaining heat will be cooled as Water reaches the radiator.
  • the stack inlet temperature control set the stack inlet target temperature T tar.in , and use the error between the current stack inlet temperature T st.in and the set target temperature T tar.in and the change rate of the temperature error as the inlet fuzzy
  • the input of the controller and the radiator air volume are the output of the fuzzy controller, which dissipates the remaining heat to the environment and finally reaches the target temperature value of the stack inlet.
  • Step S120 Determine the membership functions, fuzzy domains and fuzzy rules of the exit fuzzy controller and the entry fuzzy controller.
  • controller structure design involves the selection of membership functions, fuzzy domain and formulation of fuzzy rules.
  • the membership function forms of fuzzy controllers are diverse, the selection of fuzzy domain is also different, and the formulation of fuzzy rules also varies according to the control problem.
  • the input and output quantities of the fuzzy controller are divided into 7 fuzzy subsets using the triangular membership function, and if-then control rules are used in the fuzzy controller Take an example to illustrate. It should be understood that the ideas described are equally applicable to other types of membership functions or control rules.
  • A(x) is a function, called the function of A. membership function.
  • A(x) is a function, called the function of A. membership function.
  • the membership shape of the fuzzy subset depends on the abscissa a of the vertex of the triangle membership function, and the abscissas b and c of the base.
  • the optimization of the input and output language variables of the fuzzy controller is Optimization of the parameters a i , b i , c i (i represents different fuzzy subsets) that characterize the shape of the membership degree.
  • the input and output quantities of the fuzzy control are divided into 7 fuzzy subsets, as shown in Figure 4, namely NB (negative large), NM (negative medium), NS (negative small) , ZO (zero), PS (positive small), PM (positive middle) and PB (positive large).
  • the fuzzy universe of the stack outlet temperature error and the temperature error change rate as well as the fuzzy universe of the cooling water flow rate are designed.
  • the input and output of the fuzzy control are divided into seven fuzzy subsets, and the fuzzy domain of the stack inlet temperature error and temperature error change rate as well as the radiator wind speed are designed. Fuzzy domain of discourse.
  • Fuzzy control rules are the core of fuzzy controller and part of the knowledge base in fuzzy controller. Since the value ranges of each input and output variable are different, each basic domain of discourse is first mapped to a standardized domain of discourse with different corresponding relationships. In one embodiment, the standard domain of discussion is divided into equal parts, and then the domain of discussion is fuzzy divided to define fuzzy subsets. For example, if-then fuzzy control rules are used to formulate fuzzy rules for the controlled variables:
  • e represents the error value
  • ⁇ U represents the fuzzy controller output, see Table 1.
  • the weighted average method can be used for defuzzification.
  • Step S130 use particle swarm and genetic hybrid algorithms to optimize the membership functions of the exit fuzzy controller and the entrance fuzzy controller to achieve temperature control of the fuel cell thermal management system.
  • a particle swarm-genetic hybrid algorithm is used to optimize the membership function of the fuzzy controller, as shown in Figure 5.
  • the particle swarm-genetic hybrid algorithm it is necessary to parameterize the input and output language variables of the membership function, and then initialize each parameter, evaluate the fitness function, optimize and update the particle speed and position, as well as individual selection and crossover
  • the optimal individual in the final generation population is finally decoded, which is the optimal solution to the membership function of the optimized fuzzy controller.
  • the optimal solution is input into the fuzzy controller to improve the accuracy of the controller and keep the temperature in a smaller fluctuation range.
  • PSO Particle Swarm Optimization
  • GA Genetic Algorithm
  • the particle swarm algorithm has simple logic and fast convergence speed, but it is easy to fall into local optimality; while the genetic algorithm has strong global search ability but slow search speed. There is a strong complementarity between the two algorithms.
  • the fast convergence speed of the particle swarm algorithm is first used to perform the first stage of optimization to obtain an initial population with a certain degree of evolution. Then, the genetic algorithm performs the second stage of optimization, and finally obtains the optimal solution of the membership function, thus improving the accuracy of the fuzzy controller.
  • the control rules and the initial state of the membership function of the fuzzy controller for the stack inlet and the fuzzy controller for the stack outlet are set to be consistent. Therefore, the optimization process based on particle swarm-genetic algorithm only introduces the fuzzy controller for the stack outlet, but The optimization process is also applicable to the entry fuzzy controller.
  • the overall process of optimizing the fuzzy controller with the particle swarm-genetic hybrid algorithm includes:
  • Step S610 Randomly generate a particle population.
  • initialize the parameters in the objective function randomly generate the particle population, complete the real number encoding of the particles, and determine the fitness value range of the particles.
  • Step S620 Perform fitness evaluation.
  • Step S630 update particle speed and position.
  • the particle's own speed and position are updated according to the global optimal model loop optimization. After reaching the maximum number of iterations, the initial optimized population is output.
  • Step S640 Perform genetic operations on the initial optimized population.
  • the genetic manipulation process includes:
  • the individuals to be crossed are selected according to the size of the fitness value, and the crossover operation of the genetic algorithm is performed with the crossover probability Pc, and the partial structure of the two parent individuals is replaced and reorganized to generate a new individual;
  • Step S650 When the iteration termination condition is met, the chromosome decoding outputs the optimal parameters.
  • fuzzy control In the subsequent fuzzy control (FLC), the chromosome is assigned to the center position and width of the membership function, the control system model is run and the fitness is calculated to feed back to the particle swarm algorithm to perform fitness evaluation.
  • the overall process of fuzzy control belongs to the existing technology and will not be described again here.
  • the particle swarm optimization algorithm starts from a randomly generated population. Since the membership function needs to be optimized, the parameters to be optimized are first determined and the membership function is encoded. For example, in fuzzy control, the input and output quantities are divided into 7 fuzzy subsets, and there are 17 parameters to be optimized in the membership function, as shown in Figure 4.
  • the shape of the membership function can be determined by three points: the abscissa a of the vertex, and the abscissas b and c of the base, It is agreed that the width range of the triangle base interval is sequentially encoded as ⁇ x 1 , x 2 , x 3 , x 4 ,...x 17 ⁇ , then the center and width of the membership function can be parameterized, and the array ⁇ x 1 , x 2 , ...x 17 ⁇ is the initialized particle population, and its fitness value range is the value range [b, d] of the abscissa of the membership function, as shown in Figure 4.
  • the fitness function is a tool to produce the optimal solution.
  • the selection of the fitness function directly affects the convergence speed of the algorithm and whether the optimal solution can be found.
  • the ultimate goal of the embodiment of the present invention is to adjust the width and center position of the membership function to make temperature control more accurate. Therefore, the design of the fitness function should be as simple as possible to minimize the computational complexity.
  • the time and integrated absolute error (ITAE) performance indicators have the advantages of fast response speed and short adjustment time.
  • the ITAE performance indicator can be used as the fitness function, and the minimum fitness function value is used as the standard. Record the optimal solution of the particle itself and the optimal solution currently found by the entire population. Specifically expressed as:
  • N is the number of individuals in the population; t is time; T tar is the target temperature; T st is the current temperature of the stack.
  • the particle swarm algorithm relies on the group, so that all individuals in the group move to better-located areas based on changes in their fitness to the environment.
  • the PSO algorithm allows all particles to fly at a certain speed in the search space, and all particles dynamically adjust their flight speed based on their individual extreme values, global extreme values and other information. It is a parallel global random search algorithm.
  • the abscissa coordinate a of the vertex of the membership function, and the abscissa coordinates b and c of the bottom edge are formed into an array ⁇ x 1 , x 2 ,...x 17 ⁇ , as the initialized particle population.
  • the initialized particle population ⁇ x 1 , x 2 ,...x 17 ⁇ is the initial optimized population obtained after multiple iterations.
  • Genetic operations include selection, crossover, and mutation of individuals.
  • the purpose of selection is to select excellent individuals from the current initial optimized population so that they have the opportunity to serve as parents to reproduce for the next generation.
  • the basis for selection is that individuals with strong adaptability contribute one or more to the next generation.
  • the probability of offspring is high.
  • the roulette method determines the probability of being selected based on the fitness of the individual, that is, the selection strategy is based on the fitness ratio.
  • the probability of individual i being selected is:
  • F i and F j are the fitness values of individual i and individual j respectively; N is the number of individuals in the population.
  • the crossover operation for each individual, a new generation of individuals can be obtained by exchanging part of the chromosomes between them with the crossover probability P c .
  • the quality of the crossover operator directly affects the convergence speed of the genetic algorithm.
  • the set ⁇ x 1 , x 2 , x 3 , x 4 ...x 17 ⁇ composed of the abscissa coordinate a of the vertex of the fuzzy controller membership function and the abscissa coordinates b and c of the bottom edge forms an initial optimization after optimization by particle swarm algorithm
  • the population, chromosome is the initial optimized population.
  • the crossover operation uses the real number crossover method.
  • the method for the ⁇ -th chromosome c ⁇ and the ⁇ -th chromosome c ⁇ to cross at ⁇ is:
  • P c is the crossover probability, which is a random number in [0,1]; c ⁇ and c ⁇ are the chromosomes after the crossover operation.
  • the process is shown in Figure 7.
  • the purpose of introducing mutation is to make the genetic algorithm have local randomness. Search capabilities.
  • the particle swarm-genetic hybrid optimization algorithm of the present invention is to solve the abscissa a of the vertex of the fuzzy controller's membership function, and the abscissas b and c of the bottom. When the intersection operator is close to the optimal solution neighborhood, use This local random search capability of the mutation operator can accelerate convergence to the optimal solution.
  • the provided particle swarm-genetic hybrid algorithm optimization after parameterizing the fuzzy rules and membership functions, first uses the fast convergence speed of the particle swarm algorithm to perform the first stage of optimization, and obtains an initial population with a certain degree of evolution. Then the genetic algorithm performs the second stage of optimization, and finally obtains the optimal solution of the membership function, thereby improving the accuracy of the fuzzy controller.
  • the present invention improves the accuracy of fuel cell thermal management, with the goal of minimizing the error between the inlet and outlet temperature of the stack and the target temperature value, and uses a particle swarm-genetic hybrid optimization algorithm to subordinate the fuzzy controller
  • the center and width of the degree function are optimized.
  • the fuzzy control used has fast response speed and is suitable for the control of lagging systems.
  • the fuzzy controller optimized by the particle swarm-genetic hybrid algorithm can better resist changes in external loads, making the error between the inlet and outlet temperature and the target temperature value smaller, and can be effectively applied to the thermal management of fuel cells in high-power hybrid vehicles. , which has more advantages in accuracy and stability.
  • the present invention effectively improves the accuracy and stability of temperature control, and can be extended to temperature control of similar systems.
  • the invention may be a system, method and/or computer program product.
  • a computer program product may include a computer-readable storage medium having computer-readable program instructions thereon for causing a processor to implement various aspects of the invention.
  • Computer-readable storage media may be tangible devices that can retain and store instructions for use by an instruction execution device.
  • the computer-readable storage medium may be, for example, but not limited to, an electronic storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the above. More specific examples (non-exhaustive list) of computer-readable storage media include: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM) or Flash memory), Static Random Access Memory (SRAM), Compact Disk Read Only Memory (CD-ROM), Digital Versatile Disk (DVD), Memory Stick, Floppy Disk, Mechanical Coding Device, such as a printer with instructions stored on it.
  • RAM random access memory
  • ROM read-only memory
  • EPROM erasable programmable read-only memory
  • Flash memory Static Random Access Memory
  • CD-ROM Compact Disk Read Only Memory
  • DVD Digital Versatile Disk
  • Memory Stick
  • Computer-readable storage media are not to be construed as transient signals per se, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media (e.g., light pulses through fiber optic cables), or through electrical wires. transmitted electrical signals.
  • Computer-readable program instructions described herein may be downloaded from a computer-readable storage medium to various computing/processing devices, or to an external computer or external storage device over a network, such as the Internet, a local area network, a wide area network, and/or a wireless network.
  • the network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers, and/or edge servers.
  • a network adapter card or network interface in each computing/processing device receives computer-readable program instructions from the network and forwards the computer-readable program instructions for storage on a computer-readable storage medium in the respective computing/processing device .
  • Computer program instructions for performing operations of the present invention may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-related instructions, microcode, firmware instructions, state setting data, or instructions in one or more programming languages.
  • the computer-readable program instructions may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server implement.
  • the remote computer can be connected to the user's computer through any kind of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (such as an Internet service provider through the Internet). connect).
  • LAN local area network
  • WAN wide area network
  • an external computer such as an Internet service provider through the Internet. connect
  • an electronic circuit such as a programmable logic circuit, a field programmable gate array (FPGA), or a programmable logic array (PLA)
  • the electronic circuit can Computer readable program instructions are executed to implement various aspects of the invention.
  • These computer-readable program instructions may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus, thereby producing a machine that, when executed by the processor of the computer or other programmable data processing apparatus, , resulting in an apparatus that implements the functions/actions specified in one or more blocks in the flowchart and/or block diagram.
  • These computer-readable program instructions can also be stored in a computer-readable storage medium. These instructions cause the computer, programmable data processing device and/or other equipment to work in a specific manner. Therefore, the computer-readable medium storing the instructions includes An article of manufacture that includes instructions that implement aspects of the functions/acts specified in one or more blocks of the flowcharts and/or block diagrams.
  • Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other equipment, causing a series of operating steps to be performed on the computer, other programmable data processing apparatus, or other equipment to produce a computer-implemented process , thereby causing instructions executed on a computer, other programmable data processing apparatus, or other equipment to implement the functions/actions specified in one or more blocks in the flowcharts and/or block diagrams.
  • each block in the flowchart or block diagrams may represent a module, segment, or portion of instructions that embody one or more elements for implementing the specified logical function(s).
  • Executable instructions may occur out of the order noted in the figures. For example, two consecutive blocks may actually execute substantially in parallel, or they may sometimes execute in the reverse order, depending on the functionality involved.
  • each block of the block diagram and/or flowchart illustration, and combinations of blocks in the block diagram and/or flowchart illustration can be implemented by special purpose hardware-based systems that perform the specified functions or acts. , or can be implemented using a combination of specialized hardware and computer instructions. It is well known to those skilled in the art that implementation through hardware, implementation through software, and implementation through a combination of software and hardware are all equivalent.

Landscapes

  • Engineering & Computer Science (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • General Chemical & Material Sciences (AREA)
  • Electrochemistry (AREA)
  • Chemical Kinetics & Catalysis (AREA)
  • Chemical & Material Sciences (AREA)
  • Sustainable Energy (AREA)
  • Sustainable Development (AREA)
  • Manufacturing & Machinery (AREA)
  • Computing Systems (AREA)
  • Theoretical Computer Science (AREA)
  • Software Systems (AREA)
  • Medical Informatics (AREA)
  • Fuzzy Systems (AREA)
  • Evolutionary Computation (AREA)
  • Health & Medical Sciences (AREA)
  • Automation & Control Theory (AREA)
  • Artificial Intelligence (AREA)
  • Feedback Control In General (AREA)

Abstract

本发明公开一种燃料电池的热管理方法。该方法包括：针对燃料电池热管理系统，设置出口模糊控制器和入口模糊控制器，其中，燃料电池热管理系统包括水箱、冷却水泵、散热器和电堆，出口模糊控制器以当前电堆出口温度与设定的目标出口温度的误差和误差变化率作为输入量，以冷却水流量作为输出量；入口模糊控制器以当前电堆入口温度与设定的目标入口温度的误差和误差的变化率作为输入量，以散热器的风扇转速作为输出量；确定出口模糊控制器和入口模糊控制器的隶属度函数、模糊论域和模糊规则；对隶属度函数进行参数化处理，通过粒子群和遗传算法优化隶属度函数。本发明能精准控制燃料电池热管理系统中的温度。

Description

一种燃料电池的热管理方法 技术领域
本发明涉及燃料电池管理技术领域，更具体地，涉及一种燃料电池的热管理方法。
背景技术
氢能是一种来源丰富、绿色低碳、应用广泛的二次能源，对构建清洁低碳安全高效的能源体系、实现碳达峰碳中和目标，具有重要意义。随着新能源汽车迅速发展，氢燃料电池汽车以其高效、清洁等优势受到了广泛关注。质子交换膜燃料电池(Proton Exchange Membrane Fuel Cell，PEMFC)具有能量转换效率高、可低温运行、可靠性高和零排放等优点，在汽车领域的应用前景广阔。PEMFC是一种多物理场、多参数耦合的非线性复杂系统，其工作温度是影响输出性能和寿命的关键因素，工作温度过高会使液态水蒸发，出现膜干故障；温度过低则会导致阴极流道水淹，氧气无法穿过气体扩散层。PEMFC正常工作温度范围在60～80℃，然而在其运行过程中会产生大量的热量，因此需要对PEMFC进行有效的热管理。热管理不当会导致PEMFC输出电压出现不可逆的下降，加速其老化速度。
在现有技术中，PEMFC热管理按照原理可分为调整结构、相变冷却以及优化控制三种方式。改变流道结构会使燃料电池内部结构变复杂，增加温控循环冷却壳体，加大燃料电池体积。相变冷却使用的流体价格较高，不利于商业化，两种应用方法都存在缺陷。目前PEMFC热管理主要在温度模型上对冷却水流速以及风扇转速进行控制，控制方法有PI(Proportion Integration)控制、状态反馈控制、预测控制等。这些控制方法原理简单、使用方便，但存在响应速度慢、调节时间长等缺点。由于燃料电池固有的非线性特性以及参数的不确定性，以及应用在商用车上的高功率燃料电池输出性能和耐久性对电堆温度变化十分敏感的特点，使得现有控制方法的 应用具有一定难度。模糊控制响应速度快、抗干扰能力强，尤其适用于滞后系统的控制。有学者设计了模糊控制方法应用于PEMFC热管理中，通过调节风扇转速来控制PEMFC的温度，与以上控制方法的对比结果表明，模糊控制具有优越性。另外，有研究者考虑了克服外部负载的干扰，采用带积分的模糊控制器实时调节冷却水的流量，结果显示该方法可以在很小的波动情况下快速达到目标温度，将PEMFC电堆的温度控制在合理的范围内，相比传统的同类模型具有更强的鲁棒性。另一种现有方案是，利用改进的粒子群优化模糊PID控制，控制策略根据控制经验规则设定，具有鲁棒性强、响应速度快等优点。
经分析，现有的模糊控制设计主要依靠专家的经验，并且大多数方法采用阶跃负载信号的方式进行模糊控制的验证，然而氢燃料电池汽车在实际行驶中会有加速、匀速、减速等过程，工况的频繁变化会使得燃料电池温度控制更为复杂，温度可能会出现极端值，已有的模糊逻辑在将温度控制到目标值尚存在困难，并且在控制目标温度值处波动较大。
发明内容
本发明的目的是克服上述现有技术的缺陷，提供一种燃料电池的热管理方法，该方法包括以下步骤：
针对燃料电池热管理系统，设置出口模糊控制器和入口模糊控制器，其中，燃料电池热管理系统包括水箱、冷却水泵、散热器和电堆，所述出口模糊控制器以当前电堆出口温度与设定的目标出口温度的误差和误差变化率作为输入量，以冷却水流量作为输出量；所述入口模糊控制器以当前电堆入口温度与设定的目标入口温度的误差和误差的变化率作为输入量，以散热器的风扇转速作为输出量；
确定所述出口模糊控制器和所述入口模糊控制器的隶属度函数、模糊论域并设定模糊规则；
对所述隶属度函数进行参数化处理，以待优化参数作为粒子种群，优化所述入口模糊控制器和所述出口模糊控制器的隶属度函数，以控制燃料电池热管理系统当前的冷却水流量和散热器的风扇转速。
与现有技术相比，本发明的优点在于，以电堆的出入口温度和目标温度值之间的误差更小为目标，结合模糊控制规则以及隶属度函数优化的特点，通过整体性能更优的算法对模糊控制器的隶属度函数进行优化，经过优化的模糊控制器在稳定性和精确性上相对于常规模糊控制器提升效果明显，克服了粒子群优化算法容易陷入局部最优的缺陷。与现有优化算法相比，本发明具有更好的温度调节能力，与设定值的偏差更小，可以更好地抵御外部负载的扰动。
通过以下参照附图对本发明的示例性实施例的详细描述，本发明的其它特征及其优点将会变得清楚。
附图说明
被结合在说明书中并构成说明书的一部分的附图示出了本发明的实施例，并且连同其说明一起用于解释本发明的原理。
图1是现有技术的燃料电池热管理系统的示意图；
图2是根据本发明一个实施例的燃料电池的热管理方法的流程图；
图3是根据本发明一个实施例的模糊控制过程示意图；
图4是根据本发明一个实施例的三角形隶属度函数示意图；
图5是根据本发明一个实施例的基于粒子群-遗传混合算法优化模糊控制器示意图；
图6是根据本发明一个实施例的粒子群-遗传混合算法优化模糊控制流程图；
图7是根据本发明一个实施例的染色体交叉过程示意图。
具体实施方式
现在将参照附图来详细描述本发明的各种示例性实施例。应注意到：除非另外具体说明，否则在这些实施例中阐述的部件和步骤的相对布置、数字表达式和数值不限制本发明的范围。
以下对至少一个示例性实施例的描述实际上仅仅是说明性的，决不作为对本发明及其应用或使用的任何限制。
对于相关领域普通技术人员已知的技术、方法和设备可能不作详细讨论，但在适当情况下，所述技术、方法和设备应当被视为说明书的一部分。
在这里示出和讨论的所有例子中，任何具体值应被解释为仅仅是示例性的，而不是作为限制。因此，示例性实施例的其它例子可以具有不同的值。
应注意到：相似的标号和字母在下面的附图中表示类似项，因此，一旦某一项在一个附图中被定义，则在随后的附图中不需要对其进行进一步讨论。
为清楚起见，首先介绍PEMFC热管理系统，参见图1所示，该系统包括水箱、冷却水泵、散热器和电堆(即燃料电池堆)电堆的出入口分别设有温度传感器。在该系统中，产生的热量先被冷却水泵通过控制冷却水流量先带到水箱处，然后热量被带到散热器处，由散热器通过控制散热器风量，将热量排放到空气中。在本发明中，假设冷却水中的温度均匀，并将电堆出口冷却水温度作为电堆出口处的温度，将散热器的出口温度作为电堆的入口处温度。
参见图2所示，所提供的燃料电池的热管理方法包括以下步骤：
步骤S110，针对燃料电池热管理系统，以电堆为参考，设置出口模糊控制器和入口模糊控制器。
针对模糊逻辑设计，采用两个二维模糊控制器，对电堆出入口温度进行控制。根据选用的电堆，设定出口目标温度T tar.out，设定电堆入口目标温度T tar.in，采用合适的模糊控制规则设计模糊推理系统。在模糊推理后，反模糊化可采用加权平均法。
模糊控制实质上是一种非线性控制，从属于智能控制的范畴。在一个实施例中，建立两个曼达尼型的二维模糊控制器，分别称为出口模糊控制器和入口模糊控制器，用于对电堆出口温度和入口温度进行控制，模糊控制整体框架如图3所示。针对电堆出口温度控制，设定电堆出口目标温度T tar.out，设定当前电堆出口温度T st.out与目标出口温度T tar.out的误差和误差变化率作为出口模糊控制器的输入，冷却水流量作为出口模糊控制器的输出，通过冷却水，将电堆产生的热量通过冷却水先带到水箱处，达到电堆 出口目标温度值；热量散失掉一部分后，剩余热量随着冷却水到达散热器处。针对电堆入口温度控制，设定电堆入口目标温度T tar.in，将当前电堆入口温度T st.in与设定的目标温度T tar.in的误差和温度误差的变化率作为入口模糊控制器的输入，散热器风量为模糊控制器的输出，将剩余热量散发到环境中，最终达到电堆入口目标温度值。
步骤S120，确定出口模糊控制器和入口模糊控制器的隶属度函数、模糊论域和模糊规则。
关于基于模糊逻辑的控制器结构设计，涉及隶属度函数、模糊论域的选择以及模糊规则的制定等。模糊控制器的隶属度函数形式多样化，模糊论域的选择也不尽相同，模糊规则的制定也根据控制问题的情况各不相同。为了阐述本发明控制方法的基本原理，在下文的描述中，以三角形隶属度函数，模糊控制器的输入、输出量都划分为7个模糊子集，在模糊控制器中采用if-then控制规则为例进行说明。应理解的是，所描述的思想同样适用于其他类型的隶属度函数或控制规则。
1)隶属度函数的确定
若对论域U中的任一元素x，都有一个数A(x)∈[0，1]与之对应，当x在U中变动时，A(x)就是一个函数，称为A的隶属函数。隶属度A(x)越接近于1，表示x属于A的程度越高，A(x)越接近于0表示x属于A的程度越低。用取值于区间[0，1]的隶属函数A(x)表征x属于A的程度高低。例如，选用三角形隶属度函数，表示为：
Figure PCTCN2022092245-appb-000001
模糊子集的隶属度形状取决于三角形隶属度函数的顶点的横坐标a，以及底边的横坐标b和c，如图4所示，对模糊控制器输入以及输出语言变量的优化，也就是对表征隶属度形状的参数a i,b i,c i(i代表不同的模糊子集)的优化。
2)模糊论域的确定
在对电堆出口温度控制时，将模糊控制的输入、输出量都划分为7个模糊子集，如图4所示，即NB(负大)、NM(负中)、NS(负小)、ZO(零)、PS(正小)、PM(正中)和PB(正大)。根据温度控制目标，设计电堆出口温度误差和温度误差变化率的模糊论域以及冷却水流速的模糊论域。类似地，在设计电堆入口温度控制器时，将模糊控制的输入、输出量都划分为7个模糊子集，设计电堆入口温度误差和温度误差变化率的模糊论域以及散热器风速的模糊论域。
3)设计模糊规则
模糊控制规则是模糊控制器的核心，是模糊控制器中知识库的一部分。由于各输入输出变量取值范围各异，故首先将各基本论域分别以不同的对应关系，映射到一个标准化论域上。在一个实施例中，将标准论域等分离散化，然后对论域进行模糊划分，定义模糊子集。例如，采用if-then模糊控制规则，针对被控变量分别制定模糊规则：
If e＝NB and
Figure PCTCN2022092245-appb-000002
then ΔU＝PB；
If e＝NM and
Figure PCTCN2022092245-appb-000003
then ΔU＝PB；
……
If e＝PB and
Figure PCTCN2022092245-appb-000004
then ΔU＝NB；
其中，e代表误差值，
Figure PCTCN2022092245-appb-000005
代表误差的变化率，ΔU代表模糊控制器输出，参见表1。在模糊推理后，反模糊化可采用加权平均法。
表1 模糊控制规则示例
Figure PCTCN2022092245-appb-000006
步骤S130，使用粒子群和遗传混合算法优化出口模糊控制器和入口 模糊控制器的隶属度函数，以实现对燃料电池热管理系统的温度控制。
在一个实施例中，使用粒子群-遗传混合算法优化模糊控制器的隶属度函数，参见图5所示。采用粒子群-遗传混合算法对模糊控制器进行优化时，需对隶属度函数输入输出语言变量参数化，然后通过初始化各参数、评价适应度函数、优化更新粒子速度与位置以及个体的选择、交叉与变异等四个步骤，最终末代种群中的最优个体经过解码，即为优化模糊控制器隶属度函数的最优解。将最优解输入到模糊控制器中，从而提高控制器精准度，使温度处于更小的波动范围。
粒子群算法(Particle Swarm optimization，PSO)是通过模拟鸟群觅食行为而发展起来的一种基于群体协作的随机搜索算法。遗传算法(Genetic Algorithm，GA)是一种进化算法，其基本原理是仿效生物界中的“物竞天择、适者生存”的演化法则。粒子群算法逻辑简单，收敛速度快，但是容易陷入局部最优；而遗传算法全局搜索能力强，但搜索速度慢，这两种算法之间有着很强的互补性。
在本发明实施例中，首先利用粒子群算法收敛速度快的特点进行第一阶段的优化，得到一定进化程度的初始种群。然后，由遗传算法进行第二阶段的优化，最终获得隶属度函数的最优解，从而提高模糊控制器的精准度。为简单起见，设置电堆入口模糊控制器与电堆出口模糊控制器的控制规则以及隶属度函数初始状态一致，因此基于粒子群-遗传算法优化过程只介绍针对电堆出口的模糊控制器，但优化过程同样适用于入口模糊控制器。
参见图6所示，粒子群-遗传混合算法优化模糊控制器的整体过程包括：
步骤S610，随机生成粒子种群。
例如，初始化目标函数中的各参数，随机生成粒子种群，完成粒子的实数编码，确定粒子的适应值范围。
步骤S620，执行适应度评估。
通过适应度函数评估，记录粒子本身最优解和整个粒子种群目前找到的最优解。
步骤S630，更新粒子速度、位置。
例如，根据全局最优模型循环优化更新粒子自己的速度与位置，达到 最大迭代次数后，输出初始优化种群。
步骤S640，针对初始优化种群，执行遗传操作。
具体地，遗传操作过程包括：
在初始优化种群中，按照适应度值的大小，选择要交叉的个体，并以交叉概率Pc进行遗传算法的交叉操作，将两个父代个体的部分结构加以替换重组，生成新个体；
以P m为变异概率进行变异操作，辅助产生新个体加入子代种群中；
对于新一代种群重复上述的遗传操作，直到达到最大的进化代数G max或其他设定的终止条件。
步骤S650，在满足迭代终止条件时，染色体解码输出最优参数。
在后续的模糊控制(FLC)中，染色体赋值给隶属度函数的中心位置和宽度，运行控制系统模型并计算适应度，以反馈给粒子群算法执行适应度评估。模糊控制的整体过程属于现有技术，在此不再赘述。
以下将具体介绍上述过程中涉及的参数初始化、适应度评估、更新粒子的速度与位置、遗传操作(包括个体选择、交叉和变异)的具体实施例。
1)关于参数初始化
粒子群优化算法在从随机生成的种群开始。由于需要对隶属度函数进行优化，首先确定待优化的参数，并对隶属度函数进行编码。例如，在模糊控制中，将输入、输出量都划分为7个模糊子集，隶属度函数待优化的参数共有17个，如图4所示。为了更直观说明初始化参数过程，根据输入输出待优化的参数数目，以采用实数编码为例，隶属函数的形状可由3个点确定：顶点的横坐标a，以及底边的横坐标b和c，约定三角形底边区间宽度范围，顺次编码为{x 1,x 2,x 3,x 4,…x 17}，则隶属度函数的中心和宽度可以参数化，数组{x 1,x 2,…x 17}为初始化后的粒子种群，其适应值范围为隶属度函数横坐标的取值范围[b,d]，如图4。
2)关于适应度评估
在粒子群优化算法中，适应度函数是产出最优解的工具，适应度函数的选取直接影响算法的收敛速度以及能否找到最优解。本发明实施例的最终目标是调整隶属度函数的宽度和中心位置，使得温度控制更为精准，因 此适应度函数的设计应尽可能简单，使计算的复杂度最小。例如，时间和绝对误差积分(Integral Time-Weighted Absolute Error,ITAE)性能指标具有响应速度快、调节时间短等优点，可选用ITAE性能指标作为适应度函数，以最小化适应度函数值为标准，记录粒子本身最优解和整个种群目前找到的最优解。具体表示为：
Figure PCTCN2022092245-appb-000007
其中，N为种群中的个体数目；t为时间；T tar为目标温度；T st为电堆当前温度。
3)更新粒子自己的速度与位置
粒子群算法PSO是以群体为依托，使群体中所有的个体都根据对环境的适应度的变化移动到位置较好的区域。PSO算法让所有粒子都在搜索空间中以一定速度飞行，并且所有粒子都根据其个体极值、全局极值等信息对飞行速度进行动态调整，是一种并行的全局性的随机搜索算法。本发明实施例将隶属函数顶点的横坐标a，以及底边的横坐标b和c，组成数组{x 1,x 2,…x 17}，作为初始化后的粒子种群，根据PSO算法，粒子x i(i＝1,2,…m)在第k次迭代时，其按照全局优化模型更新速度和位置：
Figure PCTCN2022092245-appb-000008
Figure PCTCN2022092245-appb-000009
其中，v id是粒子i的飞行速度；x id是粒子i的位置；p id是粒子i所经历的最好位置(p best)；p gd是群体中所有粒子所经历过的最好位置(g best)；ω是惯性权重，负责调整粒子群的全局搜索和局部探索能力；c 1和c 2是加速度常数，表示将粒子拉向p best和g best的随机值；rand()是两个产生于[0,1]范围内的随机数。初始化后的粒子种群{x 1,x 2,…x 17}为经过多次迭代后，获得初始优化种群。
4)个体的选择、交叉与变异
遗传操作包括个体的选择、交叉和变异。对于选择操作，选择的目的是为了从当前初始优化种群中选出优良的个体，使它们有机会作为父代为下一代繁衍子孙，选择的依据是适应性强的个体为下一代贡献一个或多个 后代的概率大。例如，选用轮盘赌法进行选择，轮盘赌法由个体的适应度大小决定其被选择的概率，即基于适应度比例的选择策略，个体i被选中的概率为：
Figure PCTCN2022092245-appb-000010
其中，F i、F j分别为个体i和个体j的适应度值；N为种群中的个体数目。
对于交叉操作，对每个个体，以交叉概率P c交换它们之间的部分染色体，可以得到新一代个体。交叉算子的好坏直接影响到遗传算法的收敛速度的快慢。模糊控制器隶属度函数顶点的横坐标a，以及底边的横坐标b和c组成的集合{x 1,x 2,x 3,x 4…x 17}，经过粒子群算法优化后形成初始优化种群，染色体，即该初始优化种群。交叉操作采用实数交叉法，第α个染色体c α和第β个染色体c β在ξ处交叉的方法为：
Figure PCTCN2022092245-appb-000011
其中，P c为交叉概率，是[0,1]的随机数；c αξ、c βξ是交叉操作后的染色体，过程如图7所示。
对于变异操作，分别用符合某一范围内均匀分布的随机数，以变异概率P m来替换个体编码串中各个基因座上的原有基因值，引入变异的目的是使遗传算法具有局部的随机搜索能力。本发明的粒子群-遗传混合优化算法是为了求解模糊控制器隶属度函数顶点的横坐标a，以及底边的横坐标b和c，当通过交叉算子已接近最优解邻域时，利用变异算子的这种局部随机搜索能力可以加速向最优解收敛。
当迭代结束或者进化没有新的变化生成后，即获得最大进化代数G max，证明模糊控制隶属度函数优化结束。由于粒子群-遗传混合算法优化模糊控制器的过程比较复杂，在线优化具有一定困难，目前这方面的研究基本上都采用离线方式，即在仿真系统中取得理想效果后再拷贝入实际的模糊控制器。本发明实施例采用的也是这种离线优化方式，解码后将优化结果输入到模糊控制器中，比较在相同工作条件下优化后的模糊控制器与未经优化的模糊控制器对燃料电池出入口温度控制的效果，优化后的模糊控制相比未优化的有更高的温度控制精度。
综上，所提供的粒子群-遗传混合算法优化，将模糊规则以及隶属度函数参数化后，先利用粒子群算法收敛速度快的特点进行第一阶段的优化，得到一定进化程度的初始种群，然后由遗传算法进行第二阶段的优化，最终获得隶属度函数的最优解，从而提高模糊控制器的精准度。
综上所述，本发明针对燃料电池热管理的精度进行改进，以电堆的出入口温度和目标温度值之间的误差更小为目标，通过粒子群-遗传混合优化算法对模糊控制器的隶属度函数的中心和宽度进行优化。所采用的模糊控制响应速度快，适用于滞后系统的控制。粒子群-遗传混合算法优化后的模糊控制器能够更好地抵御外部负载的变化，使得出入口温度和目标温度值之间的误差更小，可以有效的应用到大功率混合动力汽车燃料电池热管理，在精准度和稳定性更具优势。经过计算机仿真验证，本发明有效提升了温度控制精确度和稳定度，并且可推广到类似系统的温度控制。
本发明可以是系统、方法和/或计算机程序产品。计算机程序产品可以包括计算机可读存储介质，其上载有用于使处理器实现本发明的各个方面的计算机可读程序指令。
计算机可读存储介质可以是可以保持和存储由指令执行设备使用的指令的有形设备。计算机可读存储介质例如可以是但不限于电存储设备、磁存储设备、光存储设备、电磁存储设备、半导体存储设备或者上述的任意合适的组合。计算机可读存储介质的更具体的例子(非穷举的列表)包括：便携式计算机盘、硬盘、随机存取存储器(RAM)、只读存储器(ROM)、可擦式可编程只读存储器(EPROM或闪存)、静态随机存取存储器(SRAM)、便携式压缩盘只读存储器(CD-ROM)、数字多功能盘(DVD)、记忆棒、软盘、机械编码设备、例如其上存储有指令的打孔卡或凹槽内凸起结构、以及上述的任意合适的组合。这里所使用的计算机可读存储介质不被解释为瞬时信号本身，诸如无线电波或者其他自由传播的电磁波、通过波导或其他传输媒介传播的电磁波(例如，通过光纤电缆的光脉冲)、或者通过电线传输的电信号。
这里所描述的计算机可读程序指令可以从计算机可读存储介质下载到各个计算/处理设备，或者通过网络、例如因特网、局域网、广域网和/ 或无线网下载到外部计算机或外部存储设备。网络可以包括铜传输电缆、光纤传输、无线传输、路由器、防火墙、交换机、网关计算机和/或边缘服务器。每个计算/处理设备中的网络适配卡或者网络接口从网络接收计算机可读程序指令，并转发该计算机可读程序指令，以供存储在各个计算/处理设备中的计算机可读存储介质中。
用于执行本发明操作的计算机程序指令可以是汇编指令、指令集架构(ISA)指令、机器指令、机器相关指令、微代码、固件指令、状态设置数据、或者以一种或多种编程语言的任意组合编写的源代码或目标代码，所述编程语言包括面向对象的编程语言—诸如Smalltalk、C++、Python等，以及常规的过程式编程语言—诸如“C”语言或类似的编程语言。计算机可读程序指令可以完全地在用户计算机上执行、部分地在用户计算机上执行、作为一个独立的软件包执行、部分在用户计算机上部分在远程计算机上执行、或者完全在远程计算机或服务器上执行。在涉及远程计算机的情形中，远程计算机可以通过任意种类的网络—包括局域网(LAN)或广域网(WAN)—连接到用户计算机，或者，可以连接到外部计算机(例如利用因特网服务提供商来通过因特网连接)。在一些实施例中，通过利用计算机可读程序指令的状态信息来个性化定制电子电路，例如可编程逻辑电路、现场可编程门阵列(FPGA)或可编程逻辑阵列(PLA)，该电子电路可以执行计算机可读程序指令，从而实现本发明的各个方面。
这里参照根据本发明实施例的方法、装置(系统)和计算机程序产品的流程图和/或框图描述了本发明的各个方面。应当理解，流程图和/或框图的每个方框以及流程图和/或框图中各方框的组合，都可以由计算机可读程序指令实现。
这些计算机可读程序指令可以提供给通用计算机、专用计算机或其它可编程数据处理装置的处理器，从而生产出一种机器，使得这些指令在通过计算机或其它可编程数据处理装置的处理器执行时，产生了实现流程图和/或框图中的一个或多个方框中规定的功能/动作的装置。也可以把这些计算机可读程序指令存储在计算机可读存储介质中，这些指令使得计算机、可编程数据处理装置和/或其他设备以特定方式工作，从而，存储有指令的 计算机可读介质则包括一个制造品，其包括实现流程图和/或框图中的一个或多个方框中规定的功能/动作的各个方面的指令。
也可以把计算机可读程序指令加载到计算机、其它可编程数据处理装置、或其它设备上，使得在计算机、其它可编程数据处理装置或其它设备上执行一系列操作步骤，以产生计算机实现的过程，从而使得在计算机、其它可编程数据处理装置、或其它设备上执行的指令实现流程图和/或框图中的一个或多个方框中规定的功能/动作。
附图中的流程图和框图显示了根据本发明的多个实施例的系统、方法和计算机程序产品的可能实现的体系架构、功能和操作。在这点上，流程图或框图中的每个方框可以代表一个模块、程序段或指令的一部分，所述模块、程序段或指令的一部分包含一个或多个用于实现规定的逻辑功能的可执行指令。在有些作为替换的实现中，方框中所标注的功能也可以以不同于附图中所标注的顺序发生。例如，两个连续的方框实际上可以基本并行地执行，它们有时也可以按相反的顺序执行，这依所涉及的功能而定。也要注意的是，框图和/或流程图中的每个方框、以及框图和/或流程图中的方框的组合，可以用执行规定的功能或动作的专用的基于硬件的系统来实现，或者可以用专用硬件与计算机指令的组合来实现。对于本领域技术人员来说公知的是，通过硬件方式实现、通过软件方式实现以及通过软件和硬件结合的方式实现都是等价的。
以上已经描述了本发明的各实施例，上述说明是示例性的，并非穷尽性的，并且也不限于所披露的各实施例。在不偏离所说明的各实施例的范围和精神的情况下，对于本技术领域的普通技术人员来说许多修改和变更都是显而易见的。本文中所用术语的选择，旨在最好地解释各实施例的原理、实际应用或对市场中的技术改进，或者使本技术领域的其它普通技术人员能理解本文披露的各实施例。本发明的范围由所附权利要求来限定。

Claims (10)

  1. 一种燃料电池的热管理方法，包括以下步骤：
    针对燃料电池热管理系统，设置出口模糊控制器和入口模糊控制器，其中，燃料电池热管理系统包括水箱、冷却水泵、散热器和电堆，所述出口模糊控制器以当前电堆出口温度与设定的目标出口温度的误差和误差变化率作为输入量，以冷却水流量作为输出量；所述入口模糊控制器以当前电堆入口温度与设定的目标入口温度的误差和误差的变化率作为输入量，以散热器的风扇转速作为输出量；
    确定所述出口模糊控制器和所述入口模糊控制器的隶属度函数、模糊论域并设定模糊规则；
    对所述隶属度函数进行参数化处理，以待优化参数作为粒子种群，优化所述入口模糊控制器和所述出口模糊控制器的隶属度函数，以控制燃料电池热管理系统当前的冷却水流量和散热器的风扇转速。
  2. 根据权利要求1所述的方法，其特征在于，将所述隶属度函数设置为三角形隶属度函数，该三角形隶属度函数以顶点的横坐标以及底边的两个横坐标进行表征参数，并将所述出口模糊控制器和所述入口模糊控制器的输入量、输出量划分为七个模糊子集，包括负大、负中、负小、零、正小、正中和正大。
  3. 根据权利要求2所述的方法，其特征在于，所述将隶属度函数进行参数化处理，以待优化参数作为粒子种群，优化所述入口模糊控制器和所述出口模糊控制器的隶属度函数包括：
    基于所划分的七个模糊子集对所述三角形隶属度函数的定点横坐标和底边的两个横坐标进行参数化处理，作为粒子种群，并将所述三角形隶属度函数的横坐标的取值范围作为适应值；
    以最小化设定的适应度函数为标准，确定粒子本身最优解和整个粒子种群的最优解，以调整所述三角形隶属度函数的宽度和中心位置；
    根据全局最优模型循环优化更新粒子自己的速度与位置，进而输出初始优化种群；
    以所述初始优化种群作为染色体，利用遗传算法执行个体的选择、交 叉和变异操作，以实现针对所述初始优化种群的优化过程。
  4. 根据权利要求3所述的方法，其特征在于，将所述适应度函数设置为：
    minF＝∫ 1 Nt|T tar-T st|dt
    其中，N为粒子种群中的个体数目，t为时间，T tar为目标温度，T st为电堆的当前温度。
  5. 根据权利要求3所述的方法，其特征在于，所述根据全局最优模型循环优化更新粒子自己的速度与位置包括：对于粒子i，在第k次迭代时更新的速度和位置表示为：
    Figure PCTCN2022092245-appb-100001
    Figure PCTCN2022092245-appb-100002
    其中，v id是粒子i的速度，x id是粒子i的位置，p id是粒子i所经历的最好位置，p gd是群体中所有粒子所经历过的最好位置，ω是惯性权重，c 1和c 2是加速度常数，rand()是产生于[0，1]范围内的随机数，
    Figure PCTCN2022092245-appb-100003
    是粒子更新后的位置，
    Figure PCTCN2022092245-appb-100004
    是更新后的速度，
    Figure PCTCN2022092245-appb-100005
    是粒子i更新前的速度，
    Figure PCTCN2022092245-appb-100006
    是粒子i更新前的位置。
  6. 根据权利要求3所述的方法，其特征在于，对于所述利用遗传算法执行个体的选择、交叉和变异操作，交叉操作采用实数交叉法，第α个染色体c α和第β个染色体c β在ξ处交叉表示为：
    Figure PCTCN2022092245-appb-100007
    其中，P c为交叉概率，是[0，1]的随机数，等号左侧的c αξ、c βξ是交叉操作后的染色体，等号右侧的c αξ、c βξ是交叉操作前的染色体。
  7. 根据权利要求3所述的方法，其特征在于，对于所述利用遗传算法执行个体的选择、交叉和变异操作，个体的选择采用基于适应度比例的选择策略，个体i被选中的概率表示为：
    Figure PCTCN2022092245-appb-100008
    其中，F i、F j分别为个体i和个体j的适应度值，N为粒子种群中的个体数目。
  8. 根据权利要求1所述的方法，其特征在于，将所述模糊规则采用if-then模糊控制规则。
  9. 一种计算机可读存储介质，其上存储有计算机程序，其中，该计算机程序被处理器执行时实现根据权利要求1至8中任一项所述方法的步骤。
  10. 一种计算机设备，包括存储器和处理器，在所述存储器上存储有能够在处理器上运行的计算机程序，其特征在于，所述处理器执行所述计算机程序时实现权利要求1至8中任一项所述的方法的步骤。
PCT/CN2022/092245 2022-05-11 2022-05-11 一种燃料电池的热管理方法 WO2023216150A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/CN2022/092245 WO2023216150A1 (zh) 2022-05-11 2022-05-11 一种燃料电池的热管理方法

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2022/092245 WO2023216150A1 (zh) 2022-05-11 2022-05-11 一种燃料电池的热管理方法

Publications (1)

Publication Number Publication Date
WO2023216150A1 true WO2023216150A1 (zh) 2023-11-16

Family

ID=88729287

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/092245 WO2023216150A1 (zh) 2022-05-11 2022-05-11 一种燃料电池的热管理方法

Country Status (1)

Country Link
WO (1) WO2023216150A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117518780A (zh) * 2023-12-04 2024-02-06 陕西旭氢时代科技有限公司 基于仿真模型的燃料电池热电联供系统控制方法
CN117930912A (zh) * 2024-03-22 2024-04-26 山东天力科技工程有限公司 一种基于粒子群算法的电加热焙烧窑温度控制方法

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20060053794A1 (en) * 2004-09-10 2006-03-16 Regents Of The University Of Minnesota Hybrid geothermal and fuel-cell system
CN105206852A (zh) * 2015-09-30 2015-12-30 上海电机学院 质子交换膜燃料电池的温度控制方法
CN106848349A (zh) * 2017-03-10 2017-06-13 同济大学 一种分布式燃料电池热管理系统
CN108134114A (zh) * 2017-12-07 2018-06-08 电子科技大学 一种质子交换膜燃料电池温度控制方法
CN111129548A (zh) * 2019-12-27 2020-05-08 南京航空航天大学 一种改进粒子群优化模糊pid燃料电池温度控制方法
CN113078334A (zh) * 2021-03-17 2021-07-06 电子科技大学 一种兼容不同功率电堆的燃料电池温度控制系统

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20060053794A1 (en) * 2004-09-10 2006-03-16 Regents Of The University Of Minnesota Hybrid geothermal and fuel-cell system
CN105206852A (zh) * 2015-09-30 2015-12-30 上海电机学院 质子交换膜燃料电池的温度控制方法
CN106848349A (zh) * 2017-03-10 2017-06-13 同济大学 一种分布式燃料电池热管理系统
CN108134114A (zh) * 2017-12-07 2018-06-08 电子科技大学 一种质子交换膜燃料电池温度控制方法
CN111129548A (zh) * 2019-12-27 2020-05-08 南京航空航天大学 一种改进粒子群优化模糊pid燃料电池温度控制方法
CN113078334A (zh) * 2021-03-17 2021-07-06 电子科技大学 一种兼容不同功率电堆的燃料电池温度控制系统

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117518780A (zh) * 2023-12-04 2024-02-06 陕西旭氢时代科技有限公司 基于仿真模型的燃料电池热电联供系统控制方法
CN117518780B (zh) * 2023-12-04 2024-04-09 陕西旭氢时代科技有限公司 基于仿真模型的燃料电池热电联供系统控制方法
CN117930912A (zh) * 2024-03-22 2024-04-26 山东天力科技工程有限公司 一种基于粒子群算法的电加热焙烧窑温度控制方法

Similar Documents

Publication Publication Date Title
WO2023216150A1 (zh) 一种燃料电池的热管理方法
Yuan et al. Developed coyote optimization algorithm and its application to optimal parameters estimation of PEMFC model
CN111129548B (zh) 一种改进粒子群优化模糊pid燃料电池温度控制方法
CN115084598A (zh) 一种燃料电池的热管理方法
Benchouia et al. An adaptive fuzzy logic controller (AFLC) for PEMFC fuel cell
CN112736268B (zh) 一种提高sofc系统寿命的控制优化方法和系统
CN113097542A (zh) 一种基于Amesim的燃料电池空气系统建模仿真方法
CN106654319A (zh) 一种基于变异粒子群和差分进化混合算法的pemfc系统温度建模方法
CN115392139B (zh) 一种燃料电池参数辨识方法
CN106709131A (zh) 一种适用于质子交换膜燃料电池模型的参数智能优化方法
Zhang et al. Application of coRNA-GA based RBF-NN to model proton exchange membrane fuel cells
Miao et al. Metamodel based design optimization approach in promoting the performance of proton exchange membrane fuel cells
Cao et al. PSO-Stacking improved ensemble model for campus building energy consumption forecasting based on priority feature selection
CN116435557A (zh) 一种基于神经网络的燃料电池热管理方法、装置及系统
Wang et al. Fuzzy control based on IQPSO in proton-exchange membrane fuel-cell temperature system
Wang et al. Long short-term memory deep learning model for predicting the dynamic performance of automotive PEMFC system
Fu et al. An extreme gradient boosting-based thermal management strategy for proton exchange membrane fuel cell stacks
CN112886036A (zh) 一种基于改进灰狼优化的pemfc供气系统控制策略
CN117172486A (zh) 一种基于强化学习的虚拟电厂光储资源聚合调控方法
CN114488821B (zh) 燃料电池过氧比的区间经济模型预测控制方法及系统
CN116306270A (zh) 一种预测燃料电池性能及关键输入参数寻优的方法
CN115716469A (zh) 一种混合动力系统的输出功率分配控制方法
CN113222216A (zh) 冷热电负荷预测方法、装置及系统
Huang et al. A novel parameter optimisation method of hydraulic turbine regulating system based on fuzzy differential evolution algorithm and fuzzy PID controller
Hou et al. Performance prediction of automotive fuel cell stack with genetic algorithm-BP neural network

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22941108

Country of ref document: EP

Kind code of ref document: A1