Hydrogen fuel cell system reliability evaluation method based on Weibull distribution model

Abstract

The invention relates to a hydrogen fuel cell system reliability evaluation method based on a Weibull distribution model, and the method comprises the steps: collecting the state detection quantity of a hydrogen fuel cell system in real time, and carrying out the calculation to obtain a wear index and a functional age; historical failure data of the hydrogen fuel cell system are obtained, and shape parameters and scale parameters are obtained through calculation; based on the wear index, the functional age, the shape parameter and the scale parameter, a Weibull distribution reliability model is constructed and used for calculating the cumulative failure probability and the aging failure rate; and determining the failure risk and the residual life of each component in the hydrogen fuel cell system based on the accumulated failure probability and the aging failure rate in combination with the improved risk priority. Compared with the prior art, the method has the advantages that the problems of single parameter evaluation, static analysis and inadequate individuation in the prior art are solved, dynamic, intelligent and systematic reliability evaluation is realized, and scientific support can be provided for full-life-cycle management and maintenance optimization of the hydrogen fuel cell system.

Description

Technical Field

[0001] The present invention relates to the technical field of hydrogen fuel cell aging failure detection, and in particular to a hydrogen fuel cell system reliability assessment method based on a Weibull distribution model. Background Art

[0002] As an advanced energy conversion technology, hydrogen fuel cells have become one of the important directions for future energy development with their high energy conversion efficiency, zero emissions and renewable characteristics. Unlike traditional fuel technologies, hydrogen fuel cells only produce water as a byproduct during operation and do not emit harmful gases or particulate matter. Therefore, they have attracted much attention in the context of global energy transformation. Its application prospects cover transportation, energy, aerospace and other fields, especially in clean mobile power and backup power.

[0003] Although hydrogen fuel cell technology has significant advantages, its complex structure and operating conditions bring severe challenges in reliability and safety. As a core energy device, the hydrogen fuel cell system consists of multiple key components, including hydrogen storage tanks, hydrogen pipelines, control valves and proton exchange membranes. These components operate in high temperature, high pressure and chemical corrosion environments all year round. Factors such as long-term mechanical fatigue, hydrogen embrittlement and chemical aging may cause component performance degradation or even failure. Component failure not only directly affects the stability of the fuel cell system, but may also bring potential safety risks, such as hydrogen leakage or system overheating. Therefore, how to scientifically evaluate the reliability and failure risk of these key components has become one of the core issues in promoting the widespread application of hydrogen fuel cell technology.

[0004] Most existing reliability assessment technologies use a single parameter and static model. Currently, many hydrogen fuel cell reliability studies only focus on a single component or parameter (such as pressure, temperature, etc.), without considering the coupling relationship between multiple components and multiple parameters in the system. This method ignores the mutual influence between different failure mechanisms and is difficult to fully evaluate the overall reliability of the system. Evaluation based on static empirical data cannot reflect the state changes of the equipment in real time during actual operation. In other words, the static model cannot capture the dynamic process of aging, resulting in a lag in the prediction results or even failure.

[0005] In addition, existing reliability assessments are usually based on the average failure probability, while ignoring the individual characteristics of the equipment and the differences in operating conditions. This "one-size-fits-all" approach makes it difficult to accurately predict the remaining life and failure risk of a single device. Traditional methods mainly rely on manual data analysis and offline evaluation, making it difficult to monitor the health status of the system in real time, and unable to respond quickly to emergencies, which is particularly insufficient when the equipment is running for a long time or under complex operating conditions. Summary of the invention

[0006] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and to provide a hydrogen fuel cell system reliability assessment method based on the Weibull distribution model, which can realize dynamic, intelligent and systematic reliability assessment.

[0007] The object of the present invention can be achieved by the following technical solution: A hydrogen fuel cell system reliability assessment method based on Weibull distribution model, comprising the following steps:

[0008] S1. Real-time collection of status detection quantities of the hydrogen fuel cell system, and calculation of wear index and functional age;

[0009] S2. Obtain historical failure data of the hydrogen fuel cell system and calculate Weibull distribution parameters, including shape parameters and scale parameters;

[0010] S3. Based on the wear index, functional age, shape parameter and scale parameter, a Weibull distribution reliability model is constructed to calculate the cumulative failure probability and aging failure rate;

[0011] S4. Based on the cumulative failure probability and aging failure rate, combined with the improved risk priority number, determine the failure risk and remaining life of each component in the hydrogen fuel cell system.

[0012] Furthermore, the state detection quantities of the hydrogen fuel cell system in step S1 include pressure parameters, temperature parameters, electrical parameters, and gas parameters.

[0013] Furthermore, the step S1 specifically collects state detection quantities of the hydrogen fuel cell system in real time through a sensor network, and the sensor network includes pressure sensors installed in hydrogen storage tanks and hydrogen transmission pipelines, temperature sensors arranged in the fuel cell stack and coolant channels, electrical sensors embedded in the fuel cell stack module, and flow sensors placed on the gas supply path.

[0014] Furthermore, the calculation formula of the wear index in step S1 is:

[0015]

[0016] Among them, X i ′ is the normalized value of the i-th state detection quantity, X i is the original state detection data, X max and X min are the historical maximum and minimum values ​​of the i-th state detection quantity, WI is the comprehensive wear index used to quantify the aging degree of the equipment, n is the total number of state detection quantities, ω i is the weight coefficient of the i-th state detection quantity.

[0017] Furthermore, the calculation formula of the functional age in step S1 is:

[0018]

[0019] Among them, t f is the functional age, and A, B, and D are the exponential function fitting constants between the use time and wear index of the hydrogen fuel cell system.

[0020] Furthermore, the step S2 comprises the following steps:

[0021] S21. Collect and obtain historical failure data of similar systems of hydrogen fuel cell systems, including commissioning time and aging failure records;

[0022] S22, based on the historical failure data collected and acquired in step S21, calculating the discrete failure probability of the device corresponding to each age, and then calculating the corresponding cumulative failure probability;

[0023] S23, based on the calculation results of step S22, a data table is established: the first column is age, and the second column is cumulative failure probability;

[0024] S24. According to the data table, the shape parameters and scale parameters are obtained by using the least squares fitting method.

[0025] Furthermore, the discrete failure probability of the equipment corresponding to each age in step S22 is specifically the number of aging-failed equipment at that age divided by the total number of equipment in operation.

[0026] Furthermore, the specific process of step S24 is as follows:

[0027] The shape parameter β and scale parameter η are estimated using the least squares method:

[0028]

[0029] Among them, e i is the error between the tabulated data and the Weibull distribution, n is the number of data pairs in the table;

[0030] Construct an L function that is the sum of the squares of all errors:

[0031]

[0032] make and The shape parameter β and scale parameter η are obtained by solving.

[0033] Furthermore, the distribution function expression of the cumulative failure probability and aging failure rate in step S3 is:

[0034]

[0035] Among them, the cumulative failure probability F a (t f ) represents the functional age of the component t f The probability of failure before the component is completed reflects the aging process of the component; the aging failure rate λ ag (t f ) represents the functional age of the component t f The instantaneous failure risk at a certain moment reflects the time distribution trend of failure;

[0036] When the shape parameter β < 1, it means that the failure rate decreases with time, which is applicable to early failure;

[0037] When the shape parameter β = 1, the Weibull distribution degenerates into an exponential distribution, indicating a constant failure rate, which is suitable for random failures;

[0038] When the shape parameter β>1, it means that the failure rate increases with time, which is applicable to wear failure or aging failure;

[0039] The scale parameter η determines the horizontal position of the distribution, that is, the characteristic lifetime, which represents the time scale of the data.

[0040] Furthermore, the step S4 specifically calculates the risk priority number (RPN) of each component to determine the failure risk and remaining life of each component. The calculation formula of the risk priority number is:

[0041] RPN=S·O·D·λ ag (t f )

[0042] Among them, S is severity, which indicates the degree of impact of the failure on the system or personnel. The higher the score, the more serious the impact;

[0043] O is the probability of occurrence, indicating the possibility of failure, combined with the failure probability F of the Weibull distribution a (t f ) calculation, the higher the score, the more likely failure will occur, where O = Round (F a (t f )×10), when F a (t f )≤0.1, O is close to 1, indicating a low failure probability; a (t f ) ≥ 0.9, O is close to 10, indicating a high probability of failure;

[0044] D is the detection difficulty, which indicates how difficult it is for the current monitoring method to identify the failure. The higher the score, the greater the detection difficulty.

[0045] λ ag (t f ) is the aging failure rate, which indicates the instantaneous failure risk at the current time point and is used to dynamically adjust the risk priority;

[0046] The larger the RPN score, the higher the failure risk and the lower the remaining life.

[0047] Compared with the prior art, the present invention has the following advantages:

[0048] 1. The present invention can dynamically quantify the degree of system aging by collecting and analyzing various state detection quantities (including pressure, temperature, current density, gas flow, etc.) of each component in the system in real time to calculate comprehensive wear indicators and functional age. In addition, combined with the Weibull distribution and the improved risk priority number method, the failure risk and remaining life of the equipment can be comprehensively evaluated. The present invention overcomes the problems of single parameter evaluation, static analysis and insufficient personalization in the prior art, and can realize dynamic, intelligent and systematic reliability evaluation, providing scientific support for the full life cycle management and maintenance optimization of hydrogen fuel cell systems.

[0049] 2. Comprehensive evaluation of multiple parameters to improve systematicity: Different from the existing technology which usually only focuses on a single parameter (such as pressure or temperature), the present invention comprehensively considers multiple state detection quantities in the hydrogen fuel cell system (including pressure, temperature, electrical parameters, gas flow, etc.), and constructs a comprehensive wear index (WI) through normalization and weighted combination, which can comprehensively quantify the degree of aging of the equipment.

[0050] 3. Introducing functional age to achieve dynamic reliability assessment: The present invention can dynamically assess the aging process of the equipment by calculating the functional age and combining it with real-time monitoring data, thus avoiding the shortcomings of existing static assessment methods that cannot capture changes in the operating status of the equipment. Compared with static models that rely on historical data, the present invention can reflect the dynamic status of the equipment in real time, and the prediction results are more accurate.

[0051] 4. Combine Weibull distribution to optimize failure analysis: The present invention uses the Weibull distribution model to accurately describe the failure modes (such as early failure, random failure and wear failure) of hydrogen fuel cell system components using shape parameter β and scale parameter η. Compared with traditional linear or normal distribution analysis, Weibull distribution can handle multiple failure modes more flexibly and improve the ability to describe failure time distribution.

[0052] 5. Improved risk priority number (RPN), dynamic quantification of risk: This invention introduces aging failure rate and Weibull distribution failure probability into the RPN model, dynamically quantifies the instantaneous failure risk of components, and ranks risks based on severity, probability of occurrence, and difficulty of detection. Compared with the existing static RPN model, this invention can identify high-risk components in real time and dynamically adjust risk priorities, thereby improving the accuracy and timeliness of risk assessment.

[0053] 6. Higher applicability: The method of the present invention is applicable to a variety of hydrogen fuel cell system components (such as hydrogen storage tanks, hydrogen transmission pipelines, proton exchange membranes, etc.), and can flexibly adjust parameters to adapt to different operating environments and working conditions. Compared with the existing technology that is limited to specific components or a single working condition, the present invention has a wider applicability and promotion value. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 It is a schematic diagram of the method flow of the present invention;

[0055] Figure 2 It is a schematic diagram of the application process of the embodiment;

[0056] Figure 3 Schematic diagram of the calculation process of Weibull distribution parameters in the embodiment;

[0057] Figure 4 Schematic diagram of improved risk priority numbers in the embodiment. DETAILED DESCRIPTION

[0058] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0059] Example

[0060] like Figure 1 As shown, a hydrogen fuel cell system reliability assessment method based on the Weibull distribution model includes the following steps:

[0061] S1. Real-time collection of status detection quantities of the hydrogen fuel cell system, and calculation of wear index and functional age;

[0062] S2. Obtain historical failure data of the hydrogen fuel cell system and calculate Weibull distribution parameters, including shape parameters and scale parameters;

[0063] S3. Based on the wear index, functional age, shape parameter and scale parameter, a Weibull distribution reliability model is constructed to calculate the cumulative failure probability and aging failure rate;

[0064] S4. Based on the cumulative failure probability and aging failure rate, combined with the improved risk priority number, determine the failure risk and remaining life of each component in the hydrogen fuel cell system.

[0065] This embodiment applies the above solution to construct an aging failure reliability assessment model for a hydrogen fuel cell system, such as Figure 2 As shown, the main contents are:

[0066] 1) Sensors collect state detection data in real time

[0067] During the operation of the hydrogen fuel cell system, the selection of state detection quantities is crucial. This scheme selects a series of key state parameters based on the structural characteristics and main failure mechanisms of hydrogen fuel cells, covering pressure, temperature, electrical characteristics, gas flow and other indicators. Pressure parameters such as hydrogen pipeline pressure and hydrogen storage tank pressure can directly reflect the stability of hydrogen transportation, and abnormal fluctuations may lead to leakage or insufficient hydrogen supply; temperature parameters include stack operating temperature and coolant temperature, which are used to monitor the working state of the proton exchange membrane and the efficiency of the cooling system respectively; electrical parameters such as current density and voltage deviation can evaluate the performance degradation of the stack; and gas parameters such as hydrogen flow and oxygen flow are used to detect the adequacy of the reaction supply.

[0068] The above-mentioned state detection quantities are collected in real time by high-precision sensors, and the sensor network is arranged according to the importance and failure risk of the components. In the sensor network, pressure sensors are installed at key positions of hydrogen storage tanks and hydrogen pipelines to detect dynamic pressure changes; temperature sensors are arranged in the battery stack and coolant channels to capture hot spots and abnormal temperature differences; electrical sensors are embedded in the battery stack module to monitor the uniformity of voltage distribution; flow sensors are placed in the gas supply path to monitor gas supply fluctuations. The real-time data collected by the sensor network is transmitted to the central processing system through a wireless or wired network, and the sampling frequency is dynamically adjusted according to the characteristics of the detection quantity. This embodiment sets high-frequency sampling (10 times per second) for pressure and temperature parameters, and uses a lower frequency collection (1 time per minute) for the slower changing sealing data.

[0069] 2) Calculation of wear index (WI)

[0070] Wear Index (WI) is a core parameter used to quantify the aging degree of hydrogen fuel cell system components. It generates a dimensionless value by comprehensively calculating multiple normalized state detection quantities according to certain weights to represent the overall wear state of the system. This scheme designs a calculation method based on weighted linear combination to dynamically reflect the aging process and operating state of the equipment.

[0071]

[0072] Among them, X i ′ is the normalized value of the i-th state detection quantity, X i is the original state detection data, Xmax and X min are the historical maximum and minimum values ​​of the parameter, respectively; WI is the comprehensive wear index, which is used to quantify the aging degree of the equipment; n is the total number of state detection quantities, ω i is the weight coefficient of the i-th state detection quantity.

[0073] 3) Calculation of functional age

[0074] The risk assessment model and failure factor of hydrogen fuel cell system can be expressed by exponential function. The failure factor reflects the current status of the equipment, and the quality of this indicator can directly determine the current reliability level of the system.

[0075] Combined with the actual engineering needs, it can be assumed that the use time (i.e. age) of the hydrogen fuel emergency power supply system and the wear index (obtained by normalization of the detection quantity) approximately satisfy the exponential function relationship, that is,

[0076]

[0077] Where FF is the failure factor, t f is the functional age. A, B, and D are fitting constants. These three constants can be calculated using the historically recorded detection values ​​of similar hydrogen fuel power system components through nonlinear parameter estimation methods. After obtaining these parameters, for a specific system, the age corresponding to the time point can be calculated by the following formula based on the wear index calculated based on the detection value at the current or future time point:

[0078]

[0079] The wear index of a component can represent its actual aging condition, and the calculated age is generally different from its actual age (which can be called natural age). If the calculated age is greater than its natural age, it means that it has aged more seriously than its natural age. On the contrary, if the calculated age is less than its natural age, it means that it has aged slower than its natural age. Therefore, the age calculated by this formula can be called its functional age. Functional age is an indicator that evaluates the "real" degree of aging based on the current state of the equipment, rather than relying solely on its natural age (the time since it was put into use). Functional age takes into account the changes in the state of the equipment during actual operation and provides a more accurate aging assessment.

[0080] 4) Calculation of shape parameter β and scale parameter η

[0081] In reliability engineering, understanding and predicting the life distribution of a system or component is crucial to ensuring system safety and optimizing maintenance strategies. Traditional normal distribution and exponential distribution have limitations when dealing with certain life data and cannot accurately describe different types of failure behaviors. The Weibull distribution provides greater flexibility to describe various types of failure modes by introducing shape parameters and scale parameters. The shape parameter β and scale parameter η of the hydrogen fuel cell system are expressed as follows: Figure 3 The steps shown calculate:

[0082] ① Collect the commissioning time and aging failure records of similar systems.

[0083] ② Calculate the discrete failure probability of the equipment corresponding to each age, and then calculate the corresponding cumulative failure probability, where the discrete failure probability corresponding to each age is equal to the number of aging-failed equipment at that age divided by the total number of equipment in operation.

[0084] ③ Establish a data table with the following two columns: the first column is age, and the second column is the cumulative failure probability.

[0085] ④ Use the least squares method to estimate the shape parameter β and scale parameter η:

[0086]

[0087] In the formula, e i is the error between the tabulated data and the Weibull distribution, and n is the number of data pairs in the table.

[0088] Construct an L function that is the sum of the squares of all errors:

[0089]

[0090] The shape parameter β and the scale parameter η can be obtained by and The solution is obtained.

[0091] 5) Establish Weibull distribution reliability model

[0092] In the failure analysis of hydrogen fuel cell systems, Weibull distribution can effectively describe and quantify the failure time distribution and failure mode of various components (such as valves, hydrogen pipelines, hydrogen storage tanks, proton exchange membranes, etc.). By combining the Weibull distribution model, the reliability and life of components can be more accurately evaluated, thereby optimizing maintenance strategies and improving system safety.

[0093] The aging failure process of the equipment satisfies the Weibull distribution, and the cumulative failure probability F a (t f ) represents the functional age of the component t fThe probability of failure before the component is completed reflects the aging process of the component; the aging failure rate λ ag (t f ) represents the functional age of the component t f The instantaneous failure risk at the moment reflects the time distribution trend of failure. The distribution function expression of the cumulative failure probability and aging failure rate of Weibull distribution is:

[0094]

[0095] In the formula, t f is the functional age, β is the shape parameter, and different values ​​of the shape parameter can represent different failure modes. When β<1, it means that the failure rate decreases with time, which is suitable for early failure; when β=1, the Weibull distribution degenerates into an exponential distribution, indicating a constant failure rate, which is suitable for random failure; when β>1, it means that the failure rate increases with time, which is suitable for wear failure or aging failure;

[0096] η is the scale parameter, which determines the horizontal position of the distribution, that is, the characteristic life span, which represents the time scale of the data.

[0097] 6) Risk management and decision support:

[0098] Risk Priority Number (RPN) is a commonly used risk assessment method that provides an intuitive reference for equipment maintenance and resource allocation by quantifying the severity, probability of occurrence, and difficulty of detection of failures. After calculating the RPN value of the equipment or component, sort them from high to low according to the RPN, giving priority to high-risk items. This solution combines the RPN method with the Weibull distribution reliability model for risk sorting and management of components in hydrogen fuel cell systems, and is particularly suitable for dynamic risk assessment and priority decision-making of complex systems. The improved risk priority number is calculated using the following formula:

[0099] RPN=S·O·D·λ ag (t f )

[0100] Severity (S): Indicates the degree of impact of the failure on the system or personnel. It is usually scored according to predefined standards (1-10 points). The higher the score, the more serious the impact.

[0101] Occurrence probability (O): indicates the possibility of failure, combined with the failure probability F of the Weibull distribution a (t f ) calculation, the higher the score, the more likely failure will occur. a (t f )×10), when F a (t f)≤0.1, O is close to 1, indicating a low failure probability; a (t f )≥0.9, O is close to 10, indicating a high probability of failure.

[0102] Detection difficulty (D): Indicates the difficulty of identifying the failure by the current monitoring means. The higher the score, the greater the detection difficulty.

[0103] Aging failure rate (λ ag (t f )): Indicates the instantaneous failure risk at the current time point, used to dynamically adjust the risk priority.

[0104] like Figure 4 As shown in the figure, the improved RPN reduces the aging failure rate λ ag (t f ) is incorporated into the risk assessment model, which can dynamically quantify the instantaneous failure risk of the equipment and combine it with the cumulative failure probability F a (t f 0 and traditional factors such as severity, probability of occurrence, and difficulty of detection, to provide a more comprehensive risk assessment. This method has significantly improved accuracy and dynamism. It can not only reflect the aging rate of components in real time, but also effectively identify high-risk components and optimize maintenance priorities. In addition, different failure modes are accurately modeled in combination with Weibull distribution parameters to ensure that the evaluation results are closer to the actual operating status of the equipment, providing a scientific basis for reliability management and maintenance decisions.

[0105] In summary, this solution is based on the failure mechanism and state detection data of the hydrogen fuel cell system. It collects key state parameters (such as pressure, temperature, electrical characteristics, gas flow, etc.) in real time through sensors to construct a comprehensive wear index (WI) to quantify the degree of equipment aging; the WI is dynamically calculated using the normalization and weighted linear combination method, and combined with the functional age model, the wear index is associated with the reliability level of the equipment through an exponential function to dynamically describe the actual aging process of the equipment, making up for the shortcomings of the traditional natural age assessment method. At the same time, the Weibull distribution model is used to accurately describe the failure behavior of the components through shape parameters and scale parameters, and the failure probability and aging failure rate are calculated, thereby realizing dynamic risk assessment and remaining life prediction. Among them, the cumulative failure probability function F a (t) represents the probability of equipment failure before a specific time t, and the aging failure rate function λ ag(t) quantifies the instantaneous failure risk. According to the value of the shape parameter, the Weibull distribution can flexibly describe various modes such as early failure, random failure and wear failure, and is an important tool for equipment reliability assessment and remaining life prediction. In addition, the improved risk priority number (RPN) method comprehensively incorporates the severity, probability of occurrence, detection difficulty and aging failure rate of failure into the assessment. By calculating the RPN value, the risks of equipment or components can be ranked, and high-risk projects can be prioritized, providing scientific support for equipment maintenance and decision-making. This solution significantly improves the dynamics and accuracy of reliability assessment, and provides a strong guarantee for the full life cycle management and safe operation of hydrogen fuel cell systems.

Claims

1. A hydrogen fuel cell system reliability assessment method based on Weibull distribution model, characterized in that: The following steps are involved: S1. Real-time collection of status detection quantities of the hydrogen fuel cell system, and calculation of wear index and functional age; S2. Obtain historical failure data of the hydrogen fuel cell system and calculate Weibull distribution parameters, including shape parameters and scale parameters; S3. Based on the wear index, functional age, shape parameter and scale parameter, a Weibull distribution reliability model is constructed to calculate the cumulative failure probability and aging failure rate; S4. Based on the cumulative failure probability and aging failure rate, combined with the improved risk priority number, determine the failure risk and remaining life of each component in the hydrogen fuel cell system.

2. A hydrogen fuel cell system reliability assessment method based on Weibull distribution model according to claim 1, characterized in that: The state detection quantities of the hydrogen fuel cell system in step S1 include pressure parameters, temperature parameters, electrical parameters, and gas parameters.

3. A hydrogen fuel cell system reliability assessment method based on Weibull distribution model according to claim 2, characterized in that: The step S1 specifically collects the status detection quantity of the hydrogen fuel cell system in real time through a sensor network, and the sensor network includes pressure sensors installed in the hydrogen storage tank and the hydrogen transmission pipeline, temperature sensors arranged in the fuel cell stack and the coolant channel, electrical sensors embedded in the fuel cell stack module, and flow sensors placed on the gas supply path.

4. A hydrogen fuel cell system reliability assessment method based on Weibull distribution model according to claim 2, characterized in that: The calculation formula of the wear index in step S1 is: Among them, X i ′ is the normalized value of the i-th state detection quantity, X i is the original state detection data, X max and X min are the historical maximum and minimum values ​​of the i-th state detection quantity, WI is the comprehensive wear index used to quantify the aging degree of the equipment, n is the total number of state detection quantities, ω i is the weight coefficient of the i-th state detection quantity.

5. A hydrogen fuel cell system reliability assessment method based on Weibull distribution model according to claim 4, characterized in that: The calculation formula of functional age in step S1 is: Among them, t f is the functional age, and A, B, and D are the exponential function fitting constants between the use time and wear index of the hydrogen fuel cell system.

6. A hydrogen fuel cell system reliability assessment method based on Weibull distribution model according to claim 5, characterized in that: The step S2 comprises the following steps: S21. Collect and obtain historical failure data of similar systems of hydrogen fuel cell systems, including commissioning time and aging failure records; S22, based on the historical failure data collected and acquired in step S21, calculating the discrete failure probability of the device corresponding to each age, and then calculating the corresponding cumulative failure probability; S23, based on the calculation results of step S22, a data table is established: the first column is age, and the second column is cumulative failure probability; S24. According to the data table, the shape parameters and scale parameters are obtained by using the least squares fitting method.

7. A hydrogen fuel cell system reliability assessment method based on Weibull distribution model according to claim 6, characterized in that: In step S22, the discrete failure probability of the equipment corresponding to each age is specifically the number of aging-failed equipment at that age divided by the total number of equipment in operation.

8. A hydrogen fuel cell system reliability assessment method based on Weibull distribution model according to claim 6, characterized in that: The specific process of step S24 is as follows: The shape parameter β and scale parameter η are estimated using the least squares method: Among them, e i is the error between the tabulated data and the Weibull distribution, n is the number of data pairs in the table; Construct an L function that is the sum of the squares of all errors: make and The shape parameter β and scale parameter η are obtained by solving.

9. A hydrogen fuel cell system reliability assessment method based on Weibull distribution model according to claim 8, characterized in that: The distribution function expression of the cumulative failure probability and aging failure rate in step S3 is: Among them, the cumulative failure probability F a (t f ) represents the functional age of the component t f The probability of failure before the component is completed reflects the aging process of the component; the aging failure rate λ ag (t f ) represents the functional age of the component t f The instantaneous failure risk at a certain moment reflects the time distribution trend of failure; When the shape parameter β < 1, it means that the failure rate decreases with time, which is applicable to early failure; When the shape parameter β = 1, the Weibull distribution degenerates into an exponential distribution, indicating a constant failure rate, which is suitable for random failures; When the shape parameter β>1, it means that the failure rate increases with time, which is applicable to wear failure or aging failure; The scale parameter η determines the horizontal position of the distribution, that is, the characteristic lifetime, which represents the time scale of the data.

10. A hydrogen fuel cell system reliability assessment method based on Weibull distribution model according to claim 9, characterized in that: The step S4 specifically calculates the risk priority number RPN of each component to determine the failure risk and remaining life of each component. The calculation formula of the risk priority number is: RPN=S·O·D·λ ag (t f ) Among them, S is severity, which indicates the degree of impact of the failure on the system or personnel. The higher the score, the more serious the impact; O is the probability of occurrence, indicating the possibility of failure, combined with the failure probability F of the Weibull distribution a (t f ) calculation, the higher the score, the more likely failure will occur, where O = Round (F a (t f )×10), when F a (t f )≤0.1, O is close to 1, indicating a low failure probability; a (t f ) ≥ 0.9, O is close to 10, indicating a high probability of failure; D is the detection difficulty, which indicates how difficult it is for the current monitoring method to identify the failure. The higher the score, the greater the detection difficulty. λ ag (t f ) is the aging failure rate, which indicates the instantaneous failure risk at the current time point and is used to dynamically adjust the risk priority; The larger the RPN score, the higher the failure risk and the lower the remaining life.

Images