Hybrid electric vehicle energy management method considering hydrogen fuel cell performance degradation

Abstract

The application discloses a hybrid electric vehicle energy management method considering hydrogen fuel cell performance degradation and relates to the technical field of automobile power source management. The method comprises the following steps: firstly, acquiring vehicle demand power and lithium battery state of charge, and constructing a hydrogen fuel cell performance degradation cost model; then, adopting a dynamic programming algorithm to globally optimize a preset working condition to obtain an optimal fuel cell output power sequence and form an optimal database. With demand power and state of charge as input and fuel cell output power as output, a clustering method is used to design membership function parameters, a fuzzy rule base is formulated, and each rule is fitted into an output function through linear regression. In runtime, the fuzzy controller calculates the applicable membership of each rule according to input, obtains an optimal power instruction through weighted average, and realizes hybrid electric vehicle energy management. According to the optimal energy distribution strategy, the application can solve the problem of additional hydrogen consumption caused by hydrogen fuel cell performance degradation loss and reduce the running cost of the whole vehicle.

Description

Technical Field

[0001] This invention relates to the field of automotive power source management technology, and in particular to a hybrid vehicle energy management method that takes into account the performance degradation of hydrogen fuel cells. Background Technology

[0002] With the development and promotion of new energy vehicles, hybrid power systems composed of multiple energy sources are one of the key development directions for reducing vehicle energy consumption and operating costs. Vehicles equipped with an engine, electric drive, and energy storage devices are called hybrid electric vehicles (HEVs); replacing the engine with a hydrogen fuel cell results in a hydrogen fuel cell hybrid electric vehicle (HFC). By rationally allocating the energy output of the hydrogen fuel cell and energy storage devices through energy management strategies, energy loss during vehicle operation can be reduced, thereby improving the efficiency of the powertrain. Therefore, advanced energy management strategies are crucial for hydrogen fuel cell hybrid electric vehicles.

[0003] Currently, hydrogen fuel cell hybrid electric vehicles widely adopt rule-based energy management strategies. These strategies rely on expert experience and struggle to ensure the hybrid system operates within its optimal efficiency range. Therefore, it is necessary to develop energy management strategies from the perspective of coordinated output from multiple energy sources, focusing on optimal energy consumption. Although the configurations and energy source types of hydrogen fuel cell hybrid systems vary, their core energy management objectives remain consistent: under different driving conditions, through coordinated control and rational power allocation of multiple energy sources, to achieve multi-objective optimization such as minimum energy consumption, minimum operating cost, minimum carbon emissions, or optimal system durability while meeting vehicle power requirements. From a methodological perspective, existing energy management strategies can be mainly divided into three categories: rule-based strategies, optimization theory-based strategies, and learning-based strategies.

[0004] Rule-based energy management strategies, at their core, analyze the efficiency curves of various components in the powertrain system and pre-set rules for allocating vehicle power or torque to reduce energy consumption. This method has been widely applied in hydrogen fuel cell vehicles. For example, a finite state machine-based strategy can switch the powertrain system's operating mode in real time according to preset rules based on key parameters such as brake / accelerator pedal opening, battery SOC (state of charge), and required torque. Other research introduces fuzzy logic control, dynamically optimizing control rules through fuzzification, rule base construction, reasoning, and defuzzification, effectively improving the system's robustness and flexibility in uncertain environments. While such strategies offer advantages such as simple structure, independence from precise system models, and ease of real-time implementation, they still have significant limitations: first, rule design heavily relies on expert experience and lacks systematic theoretical support, limiting applicability under complex and changing driving conditions; second, extensive parameter calibration is required to ensure stable and reliable control; and more importantly, their essence is based on static rule operation, making it difficult to adapt to dynamically changing driving conditions and multi-objective optimization needs, thus failing to fully exploit the energy-saving potential of hybrid power systems and achieve truly optimal system control.

[0005] However, there is currently no vehicle energy management strategy for multiple power sources in hybrid electric vehicles, which is based on optimization theory and learning. Therefore, there is still a technical problem of additional hydrogen consumption due to the performance degradation of hydrogen fuel cells. Summary of the Invention

[0006] The purpose of this invention is to provide an energy management method for hybrid electric vehicles that takes into account the performance degradation of hydrogen fuel cells, aiming to solve or improve at least one of the above-mentioned technical problems.

[0007] To achieve the above objectives, the present invention provides the following solution:

[0008] Energy management methods for hybrid vehicles that consider the performance degradation of hydrogen fuel cells include:

[0009] Acquire operating status data of hybrid electric vehicles; the operating status data includes vehicle power demand and lithium battery state of charge.

[0010] A performance degradation cost model for hydrogen fuel cells is constructed; the performance degradation cost model includes irreversible degradation costs and additional hydrogen consumption costs caused by reversible degradation.

[0011] Based on the performance degradation cost model, a dynamic programming algorithm is used to perform global optimization calculations on preset driving conditions to solve the optimal fuel cell output power sequence that minimizes the equivalent energy consumption of the whole vehicle. The vehicle demand power, lithium battery state of charge and corresponding optimal fuel cell output power at each time are stored together to form an optimal database.

[0012] Using the vehicle's power demand and the lithium battery's state of charge as input variables for the fuzzy controller, and the fuel cell's output power as the output variable, the optimal database is used to classify the input and output variables using a clustering method, and the parameters of the membership function of each input variable are designed based on the classification results.

[0013] Based on the classification results, a fuzzy rule base is established, and for each rule in the fuzzy rule base, a linear regression method is used to fit the corresponding output function based on the data belonging to the rule in the optimal database; the fuzzy rule base is used to map different sub-intervals of input variables to corresponding sub-intervals of output variables.

[0014] During actual vehicle operation, the collected vehicle power demand and lithium battery state of charge are input into the fuzzy controller. The applicable membership degree of each rule is calculated as a weight, and the output functions of all rules are weighted and averaged to obtain the optimal fuel cell output power command. The energy output of the hybrid vehicle is then allocated according to the optimal fuel cell output power command.

[0015] Optionally, the irreversible degradation cost is obtained by integrating the degradation rate affected by the load condition with the natural degradation rate; the additional hydrogen consumption cost caused by reversible degradation is characterized by the increase in the hydrogen consumption rate; and the increase in the hydrogen consumption rate is determined by the current compensation required to maintain the same output power.

[0016] Optionally, the cost function of the dynamic programming algorithm consists of the instantaneous hydrogen consumption cost, the irreversible degradation cost of the hydrogen fuel cell, and the equivalent hydrogen consumption cost of the lithium battery, and the terminal constraint is that the state of charge of the lithium battery is maintained within a preset range.

[0017] Optionally, the clustering method is the K-means clustering algorithm, which is used to divide the data in the optimal database into a preset number of clusters, calculate the center and variance of each cluster, and determine the Gaussian distribution parameters of the membership function.

[0018] Optionally, the membership function adopts the form of a Gaussian function, with the mean being the cluster center and the variance being the average distance or variance of the data within the cluster to the cluster center.

[0019] Optionally, the fuzzy rule base adopts the TS fuzzy control structure, which divides the input variables vehicle power demand and lithium battery state of charge into multiple fuzzy sub-intervals; the number and boundaries of the sub-intervals are determined based on the clustering classification results.

[0020] Optionally, the linear regression method uses a regular equation to solve for the regression coefficients; the regression coefficients are used to describe the linear relationship between the fuel cell output power and the input variables in any rule.

[0021] Optionally, when performing a weighted average of the output functions of all rules, the weight of each rule is taken as the membership value of the rule to the current input variable, and the final output fuel cell power command is used to control the power distribution between the hydrogen fuel cell and the lithium battery in real time.

[0022] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0023] This invention discloses an energy management method for hybrid electric vehicles that considers the performance degradation of hydrogen fuel cells. The method includes first obtaining the vehicle's required power and the state of charge (SBC) of the lithium battery, and constructing a hydrogen fuel cell performance degradation cost model. Then, a dynamic programming algorithm is used to globally optimize preset operating conditions to obtain the optimal fuel cell output power sequence, forming an optimal database. Using the required power and SBC as inputs and the fuel cell output power as outputs, a membership function parameter is designed using a clustering method, and a fuzzy rule base is established. Each rule is fitted to the output function through linear regression. During operation, the fuzzy controller calculates the applicable membership degree of each rule based on the input, and obtains the optimal power command through weighted averaging, thus realizing energy management for the hybrid electric vehicle. Based on the optimal energy allocation strategy, this invention can solve the problem of additional hydrogen consumption caused by hydrogen fuel cell performance degradation, reducing the overall vehicle operating cost. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 This is a structural diagram of the hydrogen fuel cell hybrid electric vehicle in this embodiment;

[0026] Figure 2 This is a schematic diagram of the fuzzy energy management strategy in this embodiment;

[0027] Figure 3 This is a schematic diagram of the Kmeans clustering algorithm in this embodiment. Detailed Implementation

[0028] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0029] The purpose of this invention is to provide an energy management method for hybrid electric vehicles that takes into account the performance degradation of hydrogen fuel cells, aiming to solve or improve at least one of the above-mentioned technical problems.

[0030] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0031] This invention provides an energy management method for hybrid electric vehicles that takes into account the performance degradation of hydrogen fuel cells, comprising:

[0032] Step 1: Obtain the operating status data of the hybrid vehicle; the operating status data includes the vehicle's required power and the state of charge of the lithium battery.

[0033] Step 2: Construct a performance degradation cost model for hydrogen fuel cells. This model includes irreversible degradation costs and additional hydrogen consumption costs resulting from reversible degradation. The irreversible degradation cost is obtained by integrating the degradation rate under load conditions with the natural degradation rate. The additional hydrogen consumption cost resulting from reversible degradation is characterized by the increase in the hydrogen consumption rate, which is determined by the current compensation required to maintain the same output power.

[0034] Step 3: Based on the performance degradation cost model, a dynamic programming algorithm is used to perform global optimization calculations on preset driving conditions to solve for the optimal fuel cell output power sequence that minimizes the equivalent energy consumption of the entire vehicle. The vehicle's required power, lithium battery state of charge, and corresponding optimal fuel cell output power at each time point are stored together to form an optimal database. The cost function of the dynamic programming algorithm consists of the instantaneous hydrogen consumption cost, the irreversible degradation cost of the hydrogen fuel cell, and the equivalent hydrogen consumption cost of the lithium battery, with the terminal constraint being that the lithium battery state of charge is maintained within a preset range.

[0035] Step 4: Using the vehicle's power demand and the lithium battery's state of charge as input variables for the fuzzy controller, and the fuel cell's output power as the output variable, the optimal database is used to classify the input and output variables using a clustering method. Based on the classification results, the parameters of the membership function for each input variable are designed. The clustering method is the K-means clustering algorithm, which divides the data in the optimal database into a predetermined number of clusters, calculates the center and variance of each cluster, and determines the Gaussian distribution parameters of the membership function. Furthermore, the membership function adopts a Gaussian function form, with the mean being the cluster center and the variance being the average distance or variance of the data within a cluster to the cluster center.

[0036] Step 5: Based on the classification results, a fuzzy rule base is established. For each rule in the fuzzy rule base, based on the data belonging to the rule in the optimal database, a linear regression method is used to fit the corresponding output function. The fuzzy rule base is used to map different sub-intervals of input variables to corresponding sub-intervals of output variables. Specifically, the fuzzy rule base adopts a TS fuzzy control structure, dividing the input variables vehicle power demand and lithium battery state of charge into multiple fuzzy sub-intervals. The number and boundaries of the sub-intervals are determined based on the clustering classification results. Furthermore, the linear regression method uses regular equations to solve for the regression coefficients. The regression coefficients are used to describe the linear relationship between the fuel cell output power and the input variables in any rule.

[0037] Step 6: During actual vehicle operation, the collected vehicle power demand and lithium battery state of charge are input into the fuzzy controller. The applicable membership degree of each rule is calculated as a weight, and the output functions of all rules are weighted and averaged to obtain the optimal fuel cell output power command. The energy output of the hybrid vehicle is then allocated according to the optimal fuel cell output power command. Specifically, when weighting and averaging the output functions of all rules, the weight of each rule is taken as the membership degree value of the rule with respect to the current input variable. The final output fuel cell power command is used to control the power distribution between the hydrogen fuel cell and the lithium battery in real time.

[0038] Based on the above technical solution, the following embodiment is provided, in which the following method is used: Figure 1 The hydrogen fuel cell hybrid electric vehicle structure shown in the figure manages energy; the arrows in the figure indicate energy flow.

[0039] During the use of hydrogen fuel cells, lifespan is divided into irreversible degradation and reversible degradation. Irreversible degradation is affected by operating conditions such as load conditions in addition to natural degradation. Reversible degradation is mainly affected by improper system water management, pollutants, and other conditions. The irreversible performance degradation rate of hydrogen fuel cells is expressed by equation (1):

[0040] (1)

[0041] In the formula, This indicates the total irreversible performance degradation of a hydrogen fuel cell. and These represent the performance degradation rate of the hydrogen fuel cell under load conditions and the natural degradation rate, respectively. and These represent the operating time and natural degradation time, respectively. The maximum degradation performance of a hydrogen fuel cell is defined as 10%. Exceeding this value is considered a failure in the hydrogen fuel cell's lifespan. The irreversible degradation cost is then:

[0042] (2)

[0043] The reversible performance degradation of hydrogen fuel cells is related to the hydrogen consumption rate. First, we introduce the calculation method for the hydrogen consumption rate. The hydrogen consumption rate has a linear relationship with the power of the hydrogen fuel cell, as shown below:

[0044] (3)

[0045] In the formula, C represents the linear scaling factor. fc P represents the hydrogen consumption rate. fc This indicates the power output of the hydrogen fuel cell.

[0046] The reversible performance degradation of a hydrogen fuel cell manifests as a logarithmic voltage decay, as shown below:

[0047] (4)

[0048] In the formula, Indicates the time of reversible performance degradation. Indicates the percentage of voltage degradation. This is the initial voltage of the hydrogen fuel cell. Set parameters for the model. As the hydrogen fuel cell degrades, the system current needs to be increased, i.e., the hydrogen consumption rate needs to be increased, to maintain the same output power. The hydrogen consumption rate increases as performance degrades:

[0049] (5)

[0050] Based on the equivalent consumption theory, lithium batteries can be considered as additional hydrogen storage devices. They are charged by consuming hydrogen and their hydrogen consumption is reduced by discharging. The equivalent hydrogen consumption of a lithium battery is shown below:

[0051] (6)

[0052] In the formula, This indicates the power of the lithium battery. This indicates the lower calorific value of hydrogen. as well as These represent the charging efficiency, discharging efficiency, DC / DC conversion efficiency, and efficiency of the hydrogen fuel cell, respectively.

[0053] To achieve minimum energy consumption for the entire vehicle, this invention primarily employs an energy management strategy based on fuzzy rule learning, as described below. Figure 2 As shown, the fuzzy rule learning method mainly consists of establishing the optimal database, learning membership parameters, extracting rules, and determining the controller output. First, the allocation results with minimum energy consumption are obtained based on the dynamic programming algorithm, and these results are used as the optimal database for rule learning. Second, clustering methods are used to classify the input and output of the fuzzy controller, and membership functions are designed based on the classification results. Third, the data is analyzed and extracted to design a fuzzy rule base. Finally, the controller output is obtained by fitting the data using a linear regression algorithm.

[0054] This invention employs TS fuzzy control, dividing the input into preset fuzzy sub-intervals, connecting each sub-interval through a membership function, and finally outputting the weighted average of each sub-interval, as shown below:

[0055] (7)

[0056] In the formula, It is the fuzzy range of the input. It is the output parameter, x i For input parameters. This invention sets the input to the power demand of the hydrogen fuel cell vehicle and the lithium battery. The output is the power of the hydrogen fuel cell.

[0057] The membership function is a Gaussian function, in the following form:

[0058] (8)

[0059] In the formula, and These are the mean and variance, respectively.

[0060] The optimal database is established using dynamic programming algorithm. The core idea is that in the multi-level decision-making process, regardless of the initial state and initial decision, the remaining decisions must be the optimal solutions for the state formed by the initial decisions.

[0061] Consider the following optimal control, with the cost function as follows:

[0062] (9)

[0063] In the formula, For the first The state variable at time t, These are the control variables used. For transient cost function, For terminal constraints. Given the initial state point. The dynamic equation is:

[0064] (10)

[0065] In the formula, the state satisfies the constraints. Let the control sequence This minimizes the cost function formula (9). Based on the Bellman optimal control principle, it is equivalent to:

[0066] (11)

[0067] In the formula, Let U be the cost function and U be the control parameter.

[0068] Based on the above method, the energy optimization problem based on dynamic programming algorithm is defined as follows:

[0069]

[0070] (12)

[0071] In the formula, and These represent the minimum and maximum output power of the hydrogen fuel cell, respectively. and These are the minimum and maximum values ​​for lithium batteries, respectively. . Let cost function be For a long time.

[0072] To design the membership function, this invention uses the k-means clustering method to calculate the two parameters of the Gaussian function. and The core idea of ​​the K-means clustering algorithm is to first divide the data into k groups and randomly select k pairs as cluster centers. Then, calculate the distance from each object to a cluster center, and cluster each object into the nearest cluster center. The cluster centers are then recalculated, and this process is repeated until no cluster centers change or the sum of squared errors between the current and previous cluster centers is locally minimized. The K-means clustering algorithm flowchart is shown below. Figure 3 As shown.

[0073] Based on fuzzy control theory, a rule base is defined, and the power of the hydrogen fuel cell is divided into five sub-intervals. Corresponding rule sets are generated for each sub-interval, ensuring that each type of input has a corresponding output. The specific rule base is shown in Table 1.

[0074] Table 1 Rule Base

[0075]

[0076] In the learned rules, the input of the fuzzy controller is correlated with the output power of the hydrogen fuel cell. To describe the relationship, linear regression is used to linearly fit the optimal data for each rule. The linear regression equation is shown below:

[0077] (13)

[0078] In the formula, For the first The fuel cell output power corresponding to each rule, For the corresponding input of the rule The regression coefficients, The input values ​​corresponding to the rules can therefore be written in matrix form:

[0079] (14)

[0080] In the formula, Representative Rules The dataset included uses regular equations to solve for the regression coefficients, as shown below:

[0081] (15)

[0082] After solving for the regression coefficients, we can obtain the first... Output function of the rule The output of each function is compared with the membership degree of each rule. The solution for the controller can be obtained by weighted averaging:

[0083] (16)

[0084] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0085] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A hybrid electric vehicle energy management method considering the performance degradation of hydrogen fuel cells, characterized in that, include: Acquire operating status data of hybrid electric vehicles; The operating status data includes the vehicle's required power and the lithium battery's state of charge. Construct a performance degradation cost model for hydrogen fuel cells; The performance degradation cost model includes the cost of irreversible degradation and the additional hydrogen consumption cost caused by reversible degradation. Based on the performance degradation cost model, a dynamic programming algorithm is used to perform global optimization calculations on preset driving conditions to solve the optimal fuel cell output power sequence that minimizes the equivalent energy consumption of the whole vehicle. The vehicle demand power, lithium battery state of charge and corresponding optimal fuel cell output power at each time are stored together to form an optimal database. Using the vehicle's power demand and the lithium battery's state of charge as input variables for the fuzzy controller, and the fuel cell's output power as the output variable, the optimal database is used to classify the input and output variables using a clustering method, and the parameters of the membership function of each input variable are designed based on the classification results. Based on the classification results, a fuzzy rule base is established, and for each rule in the fuzzy rule base, a linear regression method is used to fit the corresponding output function based on the data belonging to the rule in the optimal database; the fuzzy rule base is used to map different sub-intervals of input variables to corresponding sub-intervals of output variables. During actual vehicle operation, the collected vehicle power demand and lithium battery state of charge are input into the fuzzy controller. The applicable membership degree of each rule is calculated as a weight, and the output functions of all rules are weighted and averaged to obtain the optimal fuel cell output power command. The energy output of the hybrid vehicle is then allocated according to the optimal fuel cell output power command.

2. The hybrid electric vehicle energy management method considering the performance degradation of hydrogen fuel cells according to claim 1, characterized in that, The irreversible degradation cost is obtained by integrating the degradation rate affected by the load condition with the natural degradation rate; the additional hydrogen consumption cost caused by reversible degradation is characterized by the increase in the hydrogen consumption rate; the increase in the hydrogen consumption rate is determined by the current compensation required to maintain the same output power.

3. The hybrid electric vehicle energy management method considering the performance degradation of hydrogen fuel cells according to claim 1, characterized in that, The cost function of the dynamic programming algorithm consists of the instantaneous hydrogen consumption cost, the irreversible degradation cost of the hydrogen fuel cell, and the equivalent hydrogen consumption cost of the lithium battery, with the terminal constraint being that the state of charge of the lithium battery is maintained within a preset range.

4. The hybrid electric vehicle energy management method considering the performance degradation of hydrogen fuel cells according to claim 1, characterized in that, The clustering method is the K-means clustering algorithm, which is used to divide the data in the optimal database into a preset number of clusters, calculate the center and variance of each cluster, and determine the Gaussian distribution parameters of the membership function.

5. The hybrid electric vehicle energy management method considering the performance degradation of hydrogen fuel cells according to claim 4, characterized in that, The membership function adopts the form of a Gaussian function, with the mean being the cluster center and the variance being the average distance or variance of the data within the cluster to the cluster center.

6. The hybrid electric vehicle energy management method considering the performance degradation of hydrogen fuel cells according to claim 1, characterized in that, The fuzzy rule base adopts the TS fuzzy control structure, which divides the input variables, vehicle power demand and lithium battery state of charge, into multiple fuzzy sub-intervals. The number and boundaries of the sub-intervals are determined based on the clustering classification results.

7. The hybrid electric vehicle energy management method considering the performance degradation of hydrogen fuel cells according to claim 1, characterized in that, The linear regression method uses a regular equation to solve for the regression coefficients; the regression coefficients are used to describe the linear relationship between the fuel cell output power and the input variables in any rule.

8. The energy management method for hybrid electric vehicles considering the performance degradation of hydrogen fuel cells according to claim 1, characterized in that, When the output function of all rules is weighted and averaged, the weight of each rule is taken as the membership degree value of the rule to the current input variable. The final output fuel cell power command is used to control the power distribution between the hydrogen fuel cell and the lithium battery in real time.

Images