Hydrogen fuel cell temperature control method for optimizing Fuzzy-PID parameters based on IBKA

Abstract

The invention discloses a hydrogen fuel cell temperature control method for optimizing Fuzzy-PID parameters based on IBKA, and relates to the technical field of fuel cell system stability, and the method comprises the following steps: optimizing population initialization by using three strategies of random initialization, Circle chaotic mapping and a reverse learning mechanism; the convergence capability of the migration stage is improved based on the Levy Flight strategy; the output boundary is optimized by using a double-mirror reflection principle; establishing a composite fitness function which comprehensively considers dynamic performance (adjustment time and overshoot) and static performance (ITAE index) of the system; according to the method, a plurality of optimization strategies are applied to improve a basic BKA algorithm, then the algorithm is applied to parameter optimization of a fuzzy controller, the control quantity of a cooling fan is output, and therefore effective control over the temperature of a fuel cell is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of fuel cell system stability, and in particular to a hydrogen fuel cell temperature control method based on IBKA optimization of Fuzzy-PID parameters. Background Art

[0002] A hydrogen fuel cell is a fuel cell that uses hydrogen as fuel. It converts chemical energy into electrical energy through a chemical reaction between hydrogen and oxygen (usually from the air), and produces water and heat as byproducts. Among them, proton exchange membrane fuel cells have the characteristics of high energy density, zero pollution to the environment, and low noise. Because of their large output current, low operating temperature, high energy efficiency, fast startup, and compact structure, they are widely used in portable power supplies, motor vehicle power supplies, and medium and small power generation systems.

[0003] For proton exchange membrane fuel cell systems that use water cooling as the main heat dissipation method, the temperature of the coolant or the operating temperature of the fuel cell stack will affect the gas transmission, water balance and electrochemical reaction activity in the fuel cell, and will also affect the mechanical properties of the material. Temperature will directly affect the performance and efficiency of the fuel cell. Within the appropriate temperature range, the reaction rate and ion conductivity of the fuel cell are high, which can provide higher output power and power conversion efficiency. Too high or too low temperatures will lead to problems such as reduced reaction rate, electrolyte membrane moisture imbalance, and obstructed redox reactions, affecting the performance and efficiency of the fuel cell. Excessive temperatures will cause the aging and corrosion of materials to increase, reducing the service life of the fuel cell; while too low temperatures may lead to difficulties in cold starting and insufficient water analysis, affecting the stability and reliability of the fuel cell.

[0004] Therefore, an effective fuel cell temperature control method, which aims to maintain the system at an appropriate temperature, is crucial to improving the efficiency of the proton exchange membrane fuel cell system, extending the life of the proton exchange membrane fuel cell system, and improving the environmental adaptability of the fuel cell system.

[0005] However, traditional PID controllers may experience delays during the temperature adjustment process, resulting in short-term high temperatures and reducing the life of the fuel cell. Due to the unique performance of fuel cells, it is difficult to apply the established control algorithm to the proton exchange membrane fuel cell model. Therefore, it is necessary to find a suitable intelligent control strategy to iteratively optimize the fuel cell PID control parameters to achieve effective temperature control. The present invention uses three strategies: random initialization, Circle chaos mapping, and reverse learning mechanism to optimize the population initialization; based on the Lévy Flight strategy, the convergence ability of the migration stage is improved; the output boundary is optimized using the principle of double mirror reflection; a composite fitness function that comprehensively considers the dynamic performance (adjustment time and overshoot) and static performance (ITAE index) of the system is established; multiple optimization strategies are used to improve the basic black kite algorithm (BKA), and then the algorithm is applied to the parameter optimization of the fuzzy controller to output the control amount of the cooling fan, thereby achieving effective control of the fuel cell temperature. Summary of the invention

[0006] The purpose of the present invention is to solve the above-mentioned problems and to provide a hydrogen fuel cell temperature control method based on IBKA optimization of Fuzzy-PID parameters.

[0007] To achieve the above object, the technical solution adopted by the present invention is:

[0008] According to one aspect of the present invention, a hydrogen fuel cell temperature control method based on IBKA optimization of Fuzzy-PID parameters is provided, including a fuel cell temperature control system and a Fuzzy-PID temperature controller, characterized in that: the PID control parameters are optimized using an improved black kite algorithm, and the specific steps are as follows:

[0009] Step 1: Parameter initialization, set the initial PID control gain of Fuzzy-PID, the number of black kite populations N, the maximum number of iterations T, and the upper limit of the solution space and lower limit ;

[0010] Step 2: When the temperature is disturbed, the controller calculates the real-time temperature deviation based on the input actual temperature and the set target temperature, and the fuzzy controller performs feedback control based on the deviation;

[0011] Step 3: Initialize the black kite population using random initialization, Circle chaos mapping and reverse learning strategy;

[0012] Step 4: Calculate the fitness value, select the leader according to the greedy principle, and record the position X of the optimal fitness value leader ;

[0013] Step 5: In the attack phase, the position of the black kite population is updated and the fitness value is calculated;

[0014] Step 6: In the migration phase, the population position of the migration phase is generated according to the Levy flight strategy, and compared with the fitness value of the attack phase, and the optimal position with the smallest fitness value is selected to update the leader;

[0015] Step 7: Check the output solution for boundaries and make adjustments if it exceeds the limit;

[0016] Step 8: Determine whether the maximum number of iterations has been reached. If so, terminate the iteration, output the current optimal solution and assign the control gain of Fuzzy-PID. The temperature controller calculates the real-time control amount of the radiator based on the input to control the temperature of the fuel cell; if not, return to re-iterate the optimization parameters.

[0017] Furthermore, the fuel cell temperature control system includes a fuel cell heat generation model, a heat dissipation model, a water heat management system structure and a radiator model.

[0018] Furthermore, the fuzzy-PID temperature controller adopts fuzzy PID design, the specific contents are as follows:

[0019] The fuzzy PID controller adjusts the PID parameters in real time according to the designed fuzzy rules, which solves the problem that the traditional PID cannot adjust the fixed value. The fuzzy control module consists of three parts, namely, fuzzification, fuzzy reasoning, and clarity;

[0020] ① Fuzzification of input and output variables: The controller inputs are e and ec, and the fuzzy domain is [-6, 6]; the outputs are ΔK p , ΔK i , ΔK d , the fuzzy domain is [-1, 1]; the fuzzy languages ​​of input and output variables are {NB, NM, NS, ZO, PS, PM, PB};

[0021] The transformation from basic domain to fuzzy domain solves the quantization factor through the following formula:

[0022]

[0023] ② Fuzzy reasoning: In order to quickly identify the fuzzy domain and improve the matching accuracy while meeting the controller performance requirements, the temperature controller design of the present invention adopts a triangular membership function to define the fuzzy subset, and its expression formula is as follows:

[0024]

[0025] ③ Output defuzzification: Use weighted average method to adjust the proportional gain value ΔK p, differential gain adjustment value ΔK d The calculation formula is as follows:

[0026]

[0027] Where M i is the error e j and error change rate Δe j The membership function, F i is the fuzzy quantized value.

[0028] Output ΔK p , ΔK i , ΔK d Then the initial value K of the PID controller p0 , K i0 , K d0 Add together to get the PID input value K of the radiator controller p , K i , K d :

[0029] .

[0030] Compared with the prior art, the present invention has the following beneficial effects:

[0031] 1. The improved Black Kite algorithm is applied to the temperature control of fuel cells, and the key control gains of the fuzzy PID are iteratively optimized, which overcomes the disadvantage that fuzzy control is highly dependent on expert experience, improves the robustness of fuel cell temperature control, ensures stable operation of the system, and extends the service life of the fuel cell.

[0032] 2. The original population of black kites is generated based on three strategies: random initialization, Circle chaos mapping, and reverse learning mechanism, and the initial population with the best fitness value is selected according to the greedy principle, thereby enhancing the diversity of the initial population, more effectively covering the solution space, improving the quality of the initial solution, and improving the search efficiency and global search capability of the algorithm.

[0033] 3. The Lévy Flight strategy is introduced to improve the convergence ability of the migration phase, so that the Black Kite algorithm can converge faster and achieve higher accuracy, and improve the ability to escape from the local optimal solution.

[0034] 4. A hybrid fitness function that considers ITAE indicators, adjustment time and overshoot is used as an indicator to evaluate the dynamic and static performance of the system. It not only focuses on how to reach the optimization goal as quickly as possible, but also pays attention to the stability and accuracy of the process. At the same time, it meets the different priority requirements of control performance indicators in different application scenarios, balances conflicting goals, and finds the best solution that best suits a specific application. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 This is the overall flow chart of the IBKA-Fuzzy-PID temperature control of the present invention;

[0036] Figure 2 This is a diagram of the overall structure of the temperature control of a proton exchange membrane fuel cell of the present invention;

[0037] Figure 3 The IBKA-Fuzzy-PID temperature controller of the present invention;

[0038] Figure 4 is the error e membership function diagram of the present invention;

[0039] Figure 5 It is the IBKA-Fuzzy-PID control flow chart of the present invention;

[0040] Figure 6 This is a model diagram of the PEMFC water thermal management system of the present invention;

[0041] Figure 7 It is the Simulink model diagram of the controller of the present invention;

[0042] Figure 8 It is the IBKA and BKA adaptation value curve diagram of the present invention. DETAILED DESCRIPTION

[0043] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the present invention is further explained below in conjunction with specific implementation methods.

[0044] The present invention provides a hydrogen fuel cell temperature control method based on IBKA optimization of Fuzzy-PID parameters. 1. Design of fuel cell temperature control system

[0045] (1) Fuel cell heat generation model

[0046] The energy generated by the chemical reaction between hydrogen and oxygen in the fuel cell is the main source of battery energy transfer, material heating and power output, assuming that all of this chemical energy is converted into electrical energy and thermodynamic energy.

[0047] The enthalpy of hydrogen is ΔH=285.8kJ / mol, so the total chemical energy Qtot produced by the hydrogen reaction per unit time can be expressed as:

[0048] (1)

[0049] In the formula, q rec is the molar mass flow rate of the reactants.

[0050] Fuel cell output power P stIt is expressed as:

[0051] (2)

[0052] Where U st is the fuel cell stack output voltage, I st is the fuel cell stack current.

[0053] From the analysis, we know that the heat generated by the fuel cell is Q gen It can be expressed as:

[0054] (3)

[0055] (2) Heat dissipation model

[0056] Assume that the hydrogen tail gas that does not participate in the reaction has no effect on the system. The heat transfer of the fuel cell system only considers the heat dissipation of the gas and generated water and the heat dissipation of the cooling water. The heat dissipation power of the fuel cell system per unit time is:

[0057] (4)

[0058] Q cool is the heat taken away by the coolant flowing through the battery stack, expressed as:

[0059] (5)

[0060] In the formula, C w is the specific heat capacity of the coolant; M w is the coolant mass; T st,out T is the coolant outlet temperature of the battery stack; st,in The temperature of coolant entering the stack;

[0061] Q gas The heat taken away by hydrogen and oxygen entering and leaving the fuel cell stack is expressed as:

[0062] (6)

[0063] In the formula, q rec,O2 is the oxygen flow rate participating in the reaction at the cathode; q rec,H2 is the hydrogen flow rate participating in the reaction at the anode; C O2 is the specific heat capacity of oxygen; C H2 is the specific heat capacity of hydrogen; C H2O is the specific heat capacity of water vapor; T atm is the ambient temperature.

[0064] Q amb It is the heat lost by the thermal radiation of the battery stack to the external environment, which can be expressed by the ratio of the difference between the battery stack temperature and the ambient temperature to the thermal resistance:

[0065] (7)

[0066] From the above analysis, it can be seen that the PEMFC heat balance equation can be expressed as:

[0067] (8)

[0068] (3) Radiator model

[0069] The heat sink is modeled and the heat dissipation of the heat sink is expressed as:

[0070] (9)

[0071] Where K is the heat dissipation coefficient of the radiator; T rad is the radiator temperature;

[0072] The heat sink temperature can be expressed as:

[0073] (10)

[0074] In the formula, is the radiator outlet temperature; is the radiator inlet temperature.

[0075] (4) Structure of water and heat management system

[0076] The water and heat management system of the proton exchange membrane fuel cell consists of temperature sensors, radiators, circulating water pumps and constant temperature water tanks. Figure 2 shown.

[0077] The system uses water cooling to control temperature. An electronic water pump drives the coolant to circulate in the system. When the coolant flows through the stack, it absorbs the heat generated by the chemical reaction, causing the temperature to rise. Subsequently, when the coolant flows through the radiator, it exchanges heat with the outside air, releasing heat, thereby reducing the coolant temperature. This process effectively regulates the temperature of the fuel cell and keeps it within the appropriate operating range.

[0078] Design of Fuzzy-PID Temperature Controller

[0079] (1) Basic PID design

[0080] For temperature control systems, the traditional basic PID control algorithm is often used. The temperature PID controller performs feedback control based on the deviation between the target set value and the output value of the controlled object, thereby eliminating the deviation and making the controller output close to the target set value.

[0081] (11)

[0082] According to proportional-integral-differential control, the output control quantity is:

[0083] (12)

[0084] Take the integral factor and differential factor as:

[0085] (13)

[0086] According to the above formula, we can deduce:

[0087] (14)

[0088] Although adjusting K p , K i , K d Parameters can partially improve the performance of the fuel cell temperature control system, but the fuel cell system itself has the characteristics of large inertia, time-varying uncertainty, nonlinearity, etc., which leads to a large overshoot of the output result of the entire temperature control system, and the system is prone to oscillation. There is a certain conflict between the overshoot and the adjustment time, which cannot be fully taken into account, and the system lacks dynamic following ability. Therefore, the present invention builds a fuzzy PID controller to solve such problems.

[0089] (2) Fuzzy PID design

[0090] Aiming at the hysteresis and nonlinearity of fuel temperature control, fuzzy PID control is adopted to compensate for the control delay to improve the control accuracy. The fuzzy PID controller adjusts the PID parameters in real time according to the designed fuzzy rules, which solves the problem that the traditional PID cannot adjust the fixed value. The fuzzy control module consists of three parts, namely fuzzification, fuzzy reasoning, and clarification.

[0091] ① Fuzzification of input and output variables: The controller inputs are e and ec, and the fuzzy domain is [-6, 6]; the outputs are ΔK p , ΔK i , ΔK d , the fuzzy domain is [-1, 1]; the fuzzy languages ​​of input and output variables are {NB, NM, NS, ZO, PS, PM, PB};

[0092] The transformation from basic domain to fuzzy domain solves the quantization factor through the following formula:

[0093] (15)

[0094] ② Fuzzy reasoning: In order to quickly identify the fuzzy domain and improve the matching accuracy while meeting the controller performance requirements, the temperature controller design of the present invention adopts a triangular membership function to define the fuzzy subset, and its expression formula is as follows:

[0095] (16)

[0096] like Figure 4 As shown, the fuzzy rules are as follows:

[0097] Table 1 Fuzzy rules

[0098]

[0099]

[0100]

[0101] ③ Output defuzzification: Use weighted average method to adjust the proportional gain value ΔK p , differential gain adjustment value ΔK d The calculation formula is as follows:

[0102] (17)

[0103] Where M i is the error e j and error change rate Δe j The membership function, F i is the fuzzy quantized value.

[0104] Output ΔK p , ΔK i , ΔK d Then the initial value K of the PID controller p0 , K i0 , K d0 Add together to get the PID input value K of the radiator controller p , K i , K d :

[0105] (18)

[0106] (3) Controller structure design

[0107] Design the controller structure according to the control principle such as Figure 3 As shown:

[0108] Improved Black Kite Algorithm to Optimize PID Control Parameters

[0109] For nonlinear, time-varying and complex fuel cell systems, although the fuzzy PID control strategy has a good control effect on the large hysteresis of temperature, the control performance is highly dependent on expert experience. During the design, it is necessary to build a set of control logic rule bases, and these rules are generally based on experience and understanding of the system. In order to reduce this dependence and improve the adaptive ability of control performance, it is necessary to develop an adaptive parameter optimization strategy so that the temperature controller can automatically adjust the gain parameters according to the real-time feedback of the system operating status to ensure the stability of the fuel cell system operation.

[0110] The black-winged kite algorithm (BKA) is a swarm intelligence optimization algorithm based on nature inspiration, which is inspired by the survival strategy of black kites. The algorithm simulates the behavior of black kites hunting prey and population migration, and has the advantages of strong adaptability, few adjustable parameters, and high convergence accuracy. The hunting behavior of black kites can be regarded as the process of finding the optimal solution within a specific range, the prey represents the global optimal solution of the algorithm, and the migration behavior is the algorithm exploring in a larger range to avoid falling into the local optimal solution and find new potential solution areas.

[0111] The process of the original black kite algorithm:

[0112] (1) Population initialization

[0113] The black-winged kite population generates an initial solution based on the random function rand, evenly distributing the initial position of each black-winged kite. The population initialization formula is:

[0114] (19)

[0115] Where i is an integer between 1 and the population size N, j is the dimension value, and BK lb and BK ub are the lower and upper bounds of the j-th dimension black-winged kite, X i，j represents the i-th black-winged kite individual in the j-th dimension, and rand is a function that generates random values ​​between [0,1].

[0116] In the initialization stage, the greedy criterion is introduced to compare the fitness values ​​of the initial population, and the black kite individual with the best initial fitness value is selected as the leader of the initial population, marked as X. leader , which means the current best position of the black-winged kite.

[0117] (20)

[0118] (twenty one)

[0119] In the formula, f(X i，j) is the fitness function, f min The optimal fitness value obtained after traversing the population, the present invention selects the minimum value as the optimal.

[0120] (2) Aggressive behavior

[0121] The black kite is a medium-sized bird of prey belonging to the family Accipitridae. It feeds on insects, small mammals, and reptiles. When hunting, the black kite hovers in the air, using its excellent eyesight to find food on the ground. Once it finds prey, it will swoop down quickly and catch it with its sharp claws. According to its hunting habits, it can be divided into two motion states: controlled hovering and swooping hunting. The position update formula is:

[0122] (twenty two)

[0123] In the formula, and They represent the positions of the i-th black-winged kite at the t-th and t+1-th times in the j-dimension respectively; r is generated by the rand function and represents a random value between 0 and 1; q is a constant with a value of 0.9; α is a nonlinear factor, and the formula is:

[0124] (twenty three)

[0125] Where t is the current iteration number; T is the maximum iteration number.

[0126] (3) Migration behavior

[0127] The migration process of the Black Kite Algorithm is essentially a process in which the leader is responsible for guiding the entire population to move in the direction of the optimal solution. The position of the leader is not fixed, but is dynamically adjusted according to the performance of the population. The leader's affiliation is determined by evaluating the relationship between the current population fitness value and the randomly generated population fitness value. When the evaluation results show that the fitness value of the current population is lower than the fitness value of the random population, this indicates that the current leader may not have sufficient guidance ability, or the current search direction is not ideal. In this case, the leader will give up its leadership position and join the migrating population to explore new potential solution spaces with other individuals. This behavior helps to increase the diversity of the population and avoid falling into the local optimal solution, thereby improving the global search ability of the algorithm. On the contrary, if the fitness value is higher than that of the random population, it means that the leader has effectively guided the population, enabling the population to find a better solution, and the leader will continue to maintain its leadership position and lead the population towards the established goal. This continuous leadership helps to consolidate the current search direction and further optimize the quality of the solution.

[0128] The migration process formula is:

[0129] (twenty four)

[0130] (25)

[0131] In the formula, is the optimal position of the black-winged kite population in the j dimension after the tth iteration; is the fitness value of the tth iteration; is the fitness value of the random position of the population in the tth iteration.

[0132] C stands for Cauchy mutation, and the formula is as follows:

[0133] (26)

[0134] C(0,1) represents the standard Cauchy distribution ( ), its probability density function is:

[0135] (27)

[0136] The corresponding cumulative distribution function is:

[0137] (28)

[0138] In order to make the BKA algorithm have better global search capability and faster convergence speed, optimization and improvement are made to its iteration.

[0139] Initializing black-winged kite population using Circle chaos mapping and reverse learning fusion strategy

[0140] By optimizing the initialization of the black kite population, the population diversity can be enhanced, thereby improving the global search ability of the algorithm and avoiding the probability of falling into the local optimal solution.

[0141] In the present invention, the population size of the BKA algorithm is set to N, strategy 1 adopts random initialization to obtain an original population a with a population size of N, adopts Circle chaotic mapping to generate an original population b with a population size of N, adopts a dynamic reverse learning mechanism to obtain an original population c with a population size of N, and the total number of black kite populations is 3N. The fitness values ​​are calculated and sorted, and the first N populations with better fitness are taken as the initial population.

[0142] (1) Random initialization

[0143] Based on the rand random function, the original black kite algorithm initialization formula is used to produce the original population X a

[0144] (2) Circle Mapping

[0145] Circle mapping has good traversal and randomness, and can improve the initial diversity of the population. The formula for generating chaotic values ​​is:

[0146] (29)

[0147] In the formula, x i is the ith chaotic value; m is 0.2, n is 0.5; mod is the remainder function, and the chaotic state value range is (0,1). The chaotic number between 0 and 1 generated by the Circle mapping is used as the random coefficient for population initialization to generate the original population X b :

[0148] (30)

[0149] (3) Reverse learning mechanism

[0150] The reverse learning strategy covers the solution space by generating reverse solutions of the population, expanding the search range, and thus improving the diversity of the population. This method helps to make up for the deficiency of falling into the local optimum and improve the convergence speed of the algorithm. a Generate reverse population X c , the specific formula is:

[0151] (31)

[0152] Through basic random initialization, Circle chaos mapping, and reverse learning mechanism, a black kite population of 3N is obtained, and the fitness values ​​are calculated and sorted. The greedy strategy is used to make full competition within the population, and the best N individuals are selected as the initialization population. This method can make the population approach the optimal solution faster, thereby improving the convergence speed of the algorithm.

[0153] Fitness function improvement

[0154] The goal of this control system design is to achieve good temperature stability, while requiring short adjustment time, small overshoot and strong anti-interference ability. The ITAE index is included in the selection of fitness function:

[0155] (32)

[0156] The adjustment time and overshoot are introduced into the objective function as indicators for evaluating the dynamic and static performance of the system. The objective function expression selected is as follows:

[0157] (33)

[0158] Where, e(t) is the temperature deviation value; T sIndicates the adjustment time of the system; M p Indicates the overshoot of the system; and is a weight coefficient used to balance the influence of adjustment time and overshoot in the objective function. This objective function comprehensively considers the dynamic performance (adjustment time and overshoot) and static performance (ITAE index) of the system to achieve good temperature stability, short adjustment time, small overshoot and strong anti-interference ability.

[0159] Introducing the Lévy Flight strategy to improve convergence capabilities during the migration phase

[0160] The Lévy flight strategy is a random walk strategy based on the Lévy distribution, which has certain advantages when applied to optimization and search problems. The core of this strategy is to use the Lévy distribution to generate random steps. The relationship between the frequent small-distance changes and occasional long-distance jumps and time t of the Lévy flight strategy can be simplified and Fourier transformed to obtain:

[0161] (34)

[0162] Lévy's search path formula s is:

[0163] (35)

[0164] s is the Lévy flight path, parameter The value range of is (0,2), and is generally 1.5; , Both are normally distributed random numbers, and the two random variables conform to the following formula:

[0165] (36)

[0166] In the above formula, , It can be calculated by the following formula:

[0167] (37)

[0168] in Represents a gamma distribution function.

[0169] The population position iteration formula after the introduction of the Lévy flight strategy is updated as follows:

[0170] (38)

[0171] Optimizing spatial boundary based on double-sided mirror reflection principle

[0172] When the solution of an individual in the population exceeds the boundary of the problem, its value is usually restricted to the upper and lower boundaries, which may cause the solutions to be concentrated near the boundary and sparsely distributed in other areas, which may reduce the diversity of solutions and reduce the algorithm's optimization ability.

[0173] To overcome this problem, the present invention adopts a double-sided mirror reflection boundary optimization method. This method compares the boundary of the decision variable to a mirror, regards l as a light beam, and the size of l represents the intensity of light. With multiple reflections of the light beam l, the final l within the boundary x The position disappears, so l x is regarded as the projection of l within the boundary. This method can effectively solve the uneven distribution problem caused by traditional boundary processing. The specific formula is as follows:

[0174] (39)

[0175] Control strategy process

[0176] The steps of using the improved Black-winged kite optimization algorithm -Fuzzy PID IBKA-Fuzzy-PID to control the temperature of the proton exchange membrane fuel cell system are as follows:

[0177] Step 1: Parameter initialization, set the initial PID control gain of Fuzzy-PID, the number of black kite populations N, the maximum number of iterations T, and the upper limit of the solution space and lower limit ;

[0178] Step 2: When the temperature is disturbed, the controller calculates the real-time temperature deviation based on the input actual temperature and the set target temperature, and the fuzzy controller performs feedback control based on the deviation;

[0179] Step 3: Initialize the black kite population using random initialization, Circle chaos mapping and reverse learning strategy;

[0180] Step 4: Calculate the fitness value, select the leader according to the greedy principle, and record the position X of the optimal fitness value leader ;

[0181] Step 5: In the attack phase, the position of the black kite population is updated and the fitness value is calculated;

[0182] Step 6: In the migration phase, the population position of the migration phase is generated according to the Levy flight strategy, and compared with the fitness value of the attack phase, and the optimal position with the smallest fitness value is selected to update the leader;

[0183] Step 7: Check the output solution for boundaries and make adjustments if it exceeds the limit;

[0184] Step 8: Determine whether the maximum number of iterations has been reached. If so, terminate the iteration, output the current optimal solution and assign the control gain of Fuzzy-PID. The temperature controller calculates the real-time control amount of the radiator based on the input and controls the temperature of the fuel cell. If not, return to re-iterate the optimization parameters.

[0185] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

[0186] In addition, it should be understood that although the present specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.

Claims

1. A hydrogen fuel cell temperature control method based on IBKA optimization of Fuzzy-PID parameters, comprising a fuel cell temperature control system and a Fuzzy-PID temperature controller, characterized in that: The improved black kite algorithm is used to optimize the PID control parameters. The specific steps are as follows: Step 1: Parameter initialization, set the initial PID control gain of Fuzzy-PID, the number of black kite populations N, the maximum number of iterations T, and the upper limit of the solution space and lower limit ; Step 2: When the temperature is disturbed, the controller calculates the real-time temperature deviation based on the input actual temperature and the set target temperature, and the fuzzy controller performs feedback control based on the deviation; Step 3: Initialize the black kite population using random initialization, Circle chaos mapping and reverse learning strategy; Step 4: Calculate the fitness value, select the leader according to the greedy principle, and record the position of the optimal fitness value ; Step 5: In the attack phase, the position of the black kite population is updated and the fitness value is calculated; Step 6: In the migration phase, the population position of the migration phase is generated according to the Levy flight strategy, and compared with the fitness value of the attack phase, and the optimal position with the smallest fitness value is selected to update the leader; Step 7: Check the output solution and make adjustments if it exceeds the limit; Step 8: Determine whether the maximum number of iterations has been reached. If so, terminate the iteration, output the current optimal solution and assign the control gain of Fuzzy-PID. The temperature controller calculates the real-time control amount of the radiator based on the input to control the temperature of the fuel cell. If not, return to re-iterate the optimization parameters.

2. A hydrogen fuel cell temperature control method based on IBKA optimization of Fuzzy-PID parameters according to claim 1, characterized in that: The fuel cell temperature control system includes a fuel cell heat generation model, a heat dissipation model, a water heat management system structure and a radiator model.

3. A hydrogen fuel cell temperature control method based on IBKA optimization of Fuzzy-PID parameters according to claim 1, characterized in that Fuzzy-PID temperature controller adopts fuzzy PID design, the specific contents are as follows: The fuzzy PID controller adjusts the PID parameters in real time according to the designed fuzzy rules, which solves the problem that the traditional PID cannot adjust the fixed value. The fuzzy control module consists of three parts, namely fuzzification, fuzzy reasoning, and clarity; ① Fuzzification of input and output variables: The controller inputs are e and ec, and the fuzzy domain is [-6, 6]; the outputs are ΔK p , ΔK i , ΔK d , the fuzzy domain is [-1, 1]; the fuzzy languages ​​of input and output variables are {NB, NM, NS, ZO, PS, PM, PB}; The transformation from basic domain to fuzzy domain solves the quantization factor through the following formula: ； ② Fuzzy reasoning: In order to quickly identify the fuzzy domain and improve the matching accuracy while meeting the controller performance requirements, the temperature controller design of the present invention adopts a triangular membership function to define the fuzzy subset, and its expression formula is as follows: ； ③ Output defuzzification: Use weighted average method to adjust the proportional gain value ΔK p , differential gain adjustment value ΔK d The calculation formula is as follows: ； Where M i is the error e j and error change rate Δe j The membership function, F i is the fuzzy quantized value; Output ΔK p , ΔK i , ΔK d Then the initial value K of the PID controller p0 , K i0 , K d0 Add together to get the PID input value K of the radiator controller p , K i , K d : 。

Images