Rail transit product reliability evaluation method based on performance detection data

Abstract

The invention discloses a rail transit product reliability assessment method based on performance detection data, and belongs to the field of rail transit product reliability assessment. The method comprises the steps of collecting performance detection data, and establishing a performance degradation model of each component by applying a degradation track method; performing parameter estimation on the established performance degradation model to obtain degradation track model parameters of each sample; according to the degradation track model parameter of each sample and a preset failure threshold value, obtaining pseudo failure life data of each component; inputting the pseudo failure life data into the life distribution model, and calculating the reliability index of each component; and inputting the reliability index of each component into the system reliability model to calculate and obtain the reliability index of the whole rail transit product or the high-level components, and completing the comprehensive evaluation of the product reliability. The method can be applied to the condition of few or zero failure, provides more information than the failure time data, and improves the reliability of the product. And information loss caused by data loss in failure time is reduced.

Description

Technical Field

[0001] The invention relates to a rail transit product reliability assessment method based on performance detection data, belonging to the technical field of rail transit product reliability assessment. Background Art

[0002] Reliability assessment is the main content of rail transit product reliability engineering. By conducting reliability assessment on rail transit products, we can measure the level of their reliability and whether they have achieved the expected design goals. It is a necessary means to quantitatively control the reliability of rail transit products. It can also verify the rationality of the design, find out its weak links, and provide a clear direction and basis for further design improvement work. It is an important way to promote the growth of rail transit product reliability. Generally, reliability assessment refers to the process of statistically analyzing, evaluating and inferring the reliability indicators of products using probability statistics methods according to their system composition structure, distribution type and related test data throughout the product life cycle. Therefore, scientific and reasonable reliability assessment is undoubtedly a need for the development of modern industry.

[0003] Due to the integration, precision, complexity and functional diversification of rail transit products, the relationship between various systems and components in the equipment has become relatively complex. In addition, the limitations of objective conditions such as manpower, material resources, financial resources and time have brought difficulties to its reliability modeling. In addition, the high manufacturing cost and strong specialization of rail transit products make it impossible to produce them in large quantities, let alone conduct full life cycle testing and corresponding destructive life testing of products. At the same time, with the advancement of science and technology, the design and manufacturing technology of rail transit products have been continuously improved and improved. With the use of new materials, the reliability of some products has become higher and higher, and the corresponding life span has also become longer and longer, which directly leads to the inability to obtain a large amount of failure data in a relatively short period of time. In this case, if the traditional reliability evaluation method is still used, the results obtained are not convincing. Therefore, the zero failure data or less failure data of rail transit products has become an important factor restricting its reliability modeling and evaluation.

[0004] Later, researchers found that in addition to the failure data collected from the daily operation of rail transit products, there are also some information and data that can reflect its reliability characteristics to some extent during the entire life cycle of the product. For example, the "performance test data" of the product during operation contains rich reliability information. As time goes by, the material strength of the product gradually decreases, and the physical properties of the corresponding components gradually degrade until failure occurs. Therefore, the performance degradation expression can be used to describe the process of the gradual reduction of the product's function. Therefore, the research method of using performance test data to analyze the reliability of products came into being, overcoming the troubles brought by zero-failure data and less data to the reliability analysis of rail transit products. In addition, for those rail transit products with high testing costs, long cycle life and high reliability, the reliability analysis based on performance test data can not only save expensive testing costs, but also provide a good maintenance plan, eliminate unnecessary preventive maintenance, reduce its maintenance costs, reduce unnecessary waste, and improve economic benefits. Therefore, reliability analysis based on performance test data is a new trend in the development of reliability, providing a new way for traditional reliability analysis, and has very important theoretical significance and great engineering practical value.

[0005] At present, when evaluating the reliability indicators of the actual operation of rail vehicles, failure data and cumulative mileage data are usually used based on exponential distribution to calculate values ​​such as the failure rate per million kilometers (FPMK) and the mean time between failures (MTBF).

[0006] Mean time between failures (MTBF) is a basic reliability parameter for repairable products. It is measured as the ratio of the total number of product life units to the total number of failed products under specified conditions and within a specified period. The mathematical expression of MTBF is: ; Where n is the number of products, T is the total number of working hours, and r is the total number of failures.

[0007] Failure rate per million kilometers (FPMK) is a form of failure rate. Failure rate refers to a basic parameter of product reliability, which is measured as the ratio of the total number of product failures to the total number of life units under specified conditions and within a specified period. Sometimes it is also called failure rate. When the product life follows an exponential distribution, the failure rate is a constant, and its reciprocal is the average life. When the life unit is a time unit, the average life can be expressed as MTBF. However, when calculating FPMK, it is also necessary to use speed v to convert time into mileage. Finally, the mathematical expression of FPMK is: ; When using traditional reliability methods to evaluate MTBF and FPMK indicators, it is necessary to collect failure data within the evaluation range, and the data must meet certain requirements. First, the sample size of the rail transit products to be analyzed cannot be too small, and generally needs to be at least 20 pieces. If the sample is too small, the results may have large deviations. Secondly, since modern rail transit products are generally more reliable and have a longer lifespan, it may take several years or even longer to collect failure data. Third, if there is no failure of the rail transit product within the evaluation range, the MTBF cannot be directly calculated because the denominator is 0, and only a one-sided confidence lower limit at a certain confidence level can be calculated.

[0008] Traditional reliability assessment theory uses life data as the analysis object. Modern rail transit products are generally more reliable and have a longer life, but life data for high-reliability products is not easy to obtain. Whether using a timed truncation test or a constant truncation test, a certain amount of life data must be obtained to achieve an effective reliability assessment. For short-life rail transit products, life data can be easily obtained during actual operation. However, for highly reliable and long-life rail transit products, it is difficult to obtain sufficient life data within a limited product operation cycle, and traditional reliability assessment methods will lose their effectiveness. Summary of the invention

[0009] The purpose of the present invention is to overcome the deficiencies in the prior art and to provide a rail transit product reliability assessment method based on performance test data, which uses performance test data instead of failure life data for reliability assessment and life prediction. It can be applied in cases where there are only a few or zero failures, provides more information than failure time data, and reduces the information loss caused by the loss of failure time data.

[0010] To achieve the above object, the present invention is implemented by adopting the following technical solutions: In a first aspect, the present invention provides a rail transit product reliability assessment method based on performance test data, comprising: Collect performance test data of various components in rail transit products at different intervals, and process and organize the performance test data to form degradation data; Based on the degradation data, the degradation trajectory method is used to establish the performance degradation model of each component; The model parameter fitting method is used to estimate the parameters of the established performance degradation model and obtain the degradation trajectory model parameters of each sample; According to the degradation trajectory model parameters of each sample and the pre-set failure threshold, the pseudo failure life data of each component is obtained; Conduct life distribution model hypothesis test on pseudo failure life data to determine the appropriate life distribution model; Calculate the reliability index of each component according to the life distribution model; The reliability index of each component is input into the pre-built system reliability model to calculate the reliability index of the entire rail transit product or high-level components, completing a comprehensive evaluation of product reliability.

[0011] Furthermore, the method also includes: before collecting performance test data of each component in the rail transit product at different equal intervals, determining the rail transit product configuration, wherein the rail transit product configuration can manage and display the name of each component in the rail transit product and related information, the subordinate relationship between each component, and the installation location of each component, and perform statistical analysis according to different dimensions included in the configuration.

[0012] Furthermore, when the degradation trajectory method is applied to establish the performance degradation model of each component, the degradation trajectory function used includes a linear function when the second-order derivative is zero, a concave function when the second-order derivative is positive, and a convex function when the second-order derivative is negative.

[0013] Furthermore, the linear function is fitted using the following modified linear models: ; ; ; In the formula x i It is the performance degradation index of the product. i For the i Test samples, t For the detection time, α i , β i is a model parameter whose value is estimated through measured performance test data.

[0014] Furthermore, when determining the failure threshold, the selected degradation indicators include at least one or more main technical indicators. When one or more of them exceed the specified value, it is determined that the rail transit product has failed due to degradation.

[0015] Furthermore, the model parameter fitting method includes a loss function method or a least squares method.

[0016] Furthermore, the reliability indicators of each component include reliability, failure probability and failure rate.

[0017] Furthermore, the life distribution model includes a Weibull distribution model, an exponential distribution model and a normal distribution model.

[0018] Furthermore, the system reliability model includes a reliability block diagram model, a fault tree analysis model and a Monte Carlo simulation model.

[0019] Furthermore, when collecting performance test data, the sample size is at least 5, and the time interval for recording the data covers the life cycle of the component, and the corresponding number of recording time points is at least 3.

[0020] Compared with the prior art, the present invention has the following beneficial effects: The present invention provides a rail transit product reliability assessment method based on performance test data. The method collects the performance test data of each sample of the lowest-level component of the configuration at different equal intervals, determines the parameters of the performance degradation model, obtains a complete degradation model, uses the performance threshold of the component, extrapolates the failure life of each component based on the performance degradation model, further determines the failure life distribution, and then uses the life distribution model to calculate the reliability index of each component. Finally, the reliability index of a high-level component or system is calculated using a system reliability model. The present invention uses performance test data instead of failure data to perform reliability assessment and life prediction, which can be applied in the case of only a few or zero failures, provides more information than failure time data, and reduces information loss caused by the loss of failure time data. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 is a flow chart of a rail transit product reliability assessment method based on performance test data provided by an embodiment of the present invention; Figure 2 is a schematic diagram of a rail transit product configuration provided by an embodiment of the present invention; Figure 3 is a performance degradation trajectory diagram of each sample of component 1 provided in an embodiment of the present invention; Figure 4 is a pseudo failure life probability diagram of a component 1 sample under a normal distribution provided by an embodiment of the present invention; Figure 5 It is a pseudo failure life probability diagram of component 1 sample under Weibull distribution provided by an embodiment of the present invention; Figure 6 is a schematic diagram of a system reliability model of system 1 provided in an embodiment of the present invention; Figure 7 It is a schematic diagram of a system reliability model of a product provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0022] The present invention will be further described below in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and cannot be used to limit the protection scope of the present invention.

[0023] Embodiment 1: This embodiment introduces a rail transit product reliability assessment method based on performance test data, including: Collect performance test data of various components in rail transit products at different intervals, and process and organize the performance test data to form degradation data; Based on the degradation data, the degradation trajectory method is used to establish the performance degradation model of each component; The model parameter fitting method is used to estimate the parameters of the established performance degradation model and obtain the degradation trajectory model parameters of each sample; According to the degradation trajectory model parameters of each sample and the pre-set failure threshold, the pseudo failure life data of each component is obtained; Input pseudo-failure life data into a pre-built life distribution model to calculate the reliability index of each component; The reliability index of each component is input into the pre-built system reliability model to calculate the reliability index of the entire rail transit product or high-level components, completing a comprehensive evaluation of product reliability.

[0024] like Figure 1 As shown, the rail transit product reliability assessment method based on performance test data provided in this embodiment specifically involves the following steps: Determine the rail transit product configuration Configuration refers to an infrastructure model that can manage and display the names of components in rail transit products, their related information, the subordinate relationships between components, and the installation locations of components, so as to carry various types of data and perform statistical analysis according to the different dimensions contained in the configuration. Configuration is a carrier of data. Due to the different business needs of the product at different stages of its life cycle, the focus of the configuration will also be different. Therefore, the business view of the configuration will also be different. What needs to be built in this method is a configuration decomposed from a functional perspective. Determine the scope of system components that need to be covered by the evaluation based on the configuration.

[0025] Determine performance parameters and failure thresholds During the on-site use of rail transit products, there should be clear definitions of what technical status is normal and what situations no longer maintain their specified functions. Failure standards are the basis for judging the normal working status (completing the specified functions) and failure status (loss of specified functions) of rail transit products. Determining performance parameters (i.e. degradation amount) and failure standards is the basic work of degradation failure. It is necessary to select the degradation amount corresponding to the performance degradation index according to the specified functions of the rail transit products. In theory, the principle for determining the failure standard of performance degradation indicators is to select several main technical indicators as the basis for measurement. After the rail transit product has been working for a period of time, when one or several of them exceed the specified value, the rail transit product will fail due to degradation.

[0026] Collect performance data, build degradation models, and determine model parameters 3.1 Collecting performance data When collecting performance data, the sample size should be at least 5, and the time interval for recording data should cover the life cycle of the component as much as possible, with the corresponding number of recording time points being at least 3. The collected performance data is further processed and collated to form degradation data.

[0027] 3.2 Establishing a degradation model The degradation trajectory method is used to establish the degradation model corresponding to the performance parameters. Generally, a trajectory function with a stable change law is selected. g ( t ), that is, a monotonically increasing or monotonically decreasing function for modeling and analysis. According to the positive or negative nature of the second-order derivative of the degradation trajectory function, there are three commonly used degradation trajectories, namely, a linear function when the second-order derivative is zero, a concave function when the second-order derivative is positive, and a convex function when the second-order derivative is negative.

[0028] Linear degradation trajectories are generally used to describe relatively simple degradation processes, such as the wear of pantograph carbon slides and train wheel rails. Convex degradation trajectories indicate that the degradation rate of a product slows down as the amount of degradation increases, such as the growth of metal cracks. Concave degradation trajectories indicate that the degradation rate of a product accelerates as the amount of degradation increases, and are generally used to describe the degradation laws of electronic product components. In fact, many convex degradation trajectories and concave degradation trajectories can also be converted into linear degradation trajectories in certain ways (such as logarithmic coordinate changes, etc.). Degradation trajectories g ( t ) Generally, the following linear models can be used for fitting: ; ; ; Various x i It is the performance degradation index of the product. i For the i Test samples, t For the detection time, α i , β i is an unknown parameter, and its value can be estimated from the measured performance test data. The degradation trajectory curve is the trajectory of the performance degradation of a single test sample relative to time. n There are samples that show degradation. n A degradation trajectory curve.

[0029] 3.3. Determine model parameters The performance degradation trajectory of each sample of the component is estimated using model parameter fitting methods such as the loss function method or the least squares method, and finally the degradation trajectory model of each sample is obtained.

[0030] Determine the life distribution model and calculate the reliability index of the lower-level components 4.1. Selecting a lifespan distribution model Using the performance parameter failure threshold determined in step 2 x thr and the performance degradation model determined in step 3 g ( t ) and parameters, extrapolate the failure life of each sample t F . t F is the time when the degradation reaches the failure threshold, that is, Eq. g ( t )= x thr The solution.

[0031] Perform distribution hypothesis test on the data set composed of failure life of each sample, and select the distribution that the pseudo-life data may obey. The failure life of high-reliability and long-life products generally obeys Weibull distribution, exponential distribution or normal distribution, etc. Probability diagrams, goodness of fit tests, etc. can be used to judge the life distribution type and select the best one.

[0032] 4.2 Calculation of component reliability index The pseudo-failure life data obtained is regarded as full life data, and the components are evaluated according to the reliability evaluation method of the selected life distribution. The required reliability indicators such as failure rate, failure probability, and reliability are obtained.

[0033] Determine system reliability model and evaluate product reliability indicators According to the working principles and failure mechanisms of products and system components, system reliability models of products and systems are established. Models include but are not limited to RBD models, FTA models, Monte Carlo simulation models, etc. Through these system reliability models, the correlation between component reliability indicators and product reliability indicators can be established. The obtained component reliability indicators are used as inputs to system reliability models such as RBD models, FTA models, Monte Carlo simulation models, etc., so as to aggregate the reliability indicators to the product level and finally obtain the reliability indicator evaluation results of rail transit products.

[0034] The method of this embodiment is a reliability assessment method for rail transit products based on performance test data. First, the rail transit product configuration is determined according to the enterprise standard, and then the performance test data of each sample of the lowest-level components of the configuration at different equal intervals (or mileage, etc., this article uses time as an example for explanation), determines the performance degradation model, estimates the parameters of the degradation model, and obtains a complete degradation model. The performance threshold of the component is used to extrapolate the pseudo-failure life of each component based on the degradation model, and the pseudo-failure life distribution is further determined. The reliability assessment model of the life distribution is then used to calculate the reliability indicators (reliability, failure probability, failure rate) of each component. Finally, the reliability block diagram (RBD) and fault tree analysis (FTA) methods are used to calculate the reliability indicators of high-level components or systems step by step.

[0035] Using performance test data instead of failure life data for reliability assessment and life prediction has the following advantages: (1) Reliability analysis based on product performance test information can save time and money; (2) For many products, degradation is a natural property. Regardless of whether failure occurs, their performance data can be monitored and a series of degradation data can be obtained; (3) Performance test data can be used in situations where there are only a few or zero failures, providing more information than failure time data, reducing the information loss caused by the loss of failure time data; (4) For products with zero failure or very few failures, effective reliability inferences can also be obtained using degradation data.

[0036] The contents involved in the above embodiment are described below in conjunction with a preferred embodiment.

[0037] Determine the rail transit product configuration like Figure 2 As shown in the figure, if a rail transit product is composed of 4 systems (system 1 to system 4), system 1 is composed of 3 components (component 1 to component 3).

[0038] Determine performance parameters and failure thresholds Determine the amount of degradation based on the product and related system functions x is the drift of the component output, that is, t i The degradation value at the moment x i For components in t i Output value at time O i and t =0 initial output value O The difference of 1, xi = O i - O 1。

[0039] According to relevant technical requirements, a component failure is defined as a degradation value greater than 10.

[0040] Collect performance data, build degradation models, and determine model parameters The measured value of the degradation amount x obtained from the performance test data of component 1 is as follows: Table 1 Measured degradation values ​​of component 1 at various time points The number of samples recorded in the above table is 5, and the number of recorded time points is 7. According to the definition of component failure when the degradation value is greater than 10, it can be seen that by 1500 hours, although the degradation values ​​of the 5 samples have an obvious upward trend over time, they have not reached the failure standard, which means that these samples of component 1 can still perform their functions normally from the perspective of reliability.

[0041] From the data in Table 1, we can get the performance degradation trajectory as follows: Figure 3 As shown; Depend on Figure 3 It can be seen that the performance degradation trajectory of each sample of this example component is closer to a straight line, so the straight line model can be selected ( ) as the performance degradation trajectory model of the component.

[0042] Next, we need to determine the two parameters of the model , Taking sample 1 as an example, sample 1 has 7 data points, and the coordinates on the degradation trajectory diagram ( t i , y i ) are: (0,0), (250,0.47), (500,0.93), (750,2.11), (1000,2.72), (1250,3.51), (1500,4.34); For the straight line model, a loss function can be constructed to solve the model parameters: ; Treat the loss function as a multivariate function and use the method of partial derivative of multivariate function to calculate the minimum value of the function. In this example, Respectively , Find the partial derivative, set the partial derivative = 0, and get: ; ; Combining the two equations, we can get: ; ; Substituting the data points into the formula, we get: .

[0043] Then the degradation trajectory model of the sample is obtained as follows: ; The calculation method of the degradation trajectory model of other samples is the same as above, and finally the degradation trajectory model parameter results of the 5 samples of this component are as follows: Table 2 Parameter results of degradation trajectory model Determine the life distribution model and calculate the reliability index of the lower-level components Taking sample 1 as an example, it is known that the performance threshold of the component is no more than 10, that is, when the performance degradation is greater than 10, the component is judged to be failed. Substituting the threshold into the degradation trajectory formula, we get: ; Solving for t, we get t=3408.93. Since the degradation trajectory function is monotonically increasing, the failure life of the sample is 3408.93.

[0044] The failure life calculation method of other samples is the same, and finally the pseudo failure life data set of all samples of the component can be obtained.

[0045] Table 3 Pseudo failure life calculation results of each sample of component 1 Generally speaking, Weibull distribution has better adaptability, while normal distribution is relatively simple to calculate. Therefore, these two distributions are used to test the distribution hypothesis of the pseudo failure life data set. The results are as follows: Figure 4 , Figure 5 As shown; Depend on Figure 4 , Figure 5 It can be seen that the failure life points of the component samples are basically distributed on a straight line in the two probability graphs. It can be considered that both distribution assumptions are basically met. The figure also shows the calculated Anderson-Darling test statistic (AD) and its corresponding significance test probability value (P value). According to relevant theories, the larger the P value, the better the fit. Considering that the P value of the normal distribution is higher, the normal distribution is selected as the life distribution model for further analysis.

[0046] The reliability index formula corresponding to the normal distribution is as follows: Failure probability: ; Reliability: ; Failure rate: ; in: t For time, μ is the mean life span, σ is the standard deviation of life span, () is the probability density function of the standard normal distribution, and Φ() is the cumulative distribution function of the standard normal distribution.

[0047] Depend on Figure 3 It can be seen that the mean life of component 1 under normal distribution is μ is 4490, the standard deviation of life span σ It is 891.6.

[0048] It can be obtained that at t=1750, the reliability of component 1 is .

[0049] System 1 consists of three different components. If any one of the components fails, then system 1 fails. Therefore, the system is considered to be a series system. Its system reliability model can be expressed by the RBD (reliability block diagram) model as follows: Figure 6 As shown; Assume that the calculations in the first four steps show that the reliability of component 1 is 0.999, the reliability of component 2 is 0.995, and the reliability of component 3 is 0.980. Then the reliability R of system 1 can be obtained by the RBD model, that is, the reliability R of system 1 is the product of the reliabilities of the three components, and the calculation result is R=0.999*0.995*0.980=0.974, and the corresponding failure probability is F=1-0.974=0.026.

[0050] Furthermore, if the system reliability model of the product can be represented by an FTA (fault tree) model with product failure as the top event, Figure 7 As shown; Assume that the above steps calculate that the failure probability of system 1 is 0.026, the failure probability of system 2 is 0.014, the failure probability of system 3 is 0.005, and the failure probability of system 4 is 0.001. According to the logical relationship of the fault tree, the failure probability of the product F prod for: ; The corresponding product reliability is 0.999999802.

[0051] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

[0052] It should be understood by those skilled in the art that the embodiments of the present disclosure may be provided as methods, systems or computer program products. Therefore, the present disclosure may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware aspects. Moreover, the present disclosure may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0053] The present disclosure is described with reference to flowcharts and / or block diagrams of methods, devices (systems) and computer program products according to embodiments of the present disclosure. It should be understood that each process and / or block in the flowchart and / or block diagram and the combination of processes and / or blocks in the flowchart and / or block diagram can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0054] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0055] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0056] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present disclosure rather than to limit its protection scope. Although the present disclosure has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that after reading the present disclosure, those skilled in the art can still make various changes, modifications or equivalent substitutions to the specific implementation methods of the invention, but these changes, modifications or equivalent substitutions are all within the protection scope of the disclosed claims to be approved.

Claims

1. A rail transit product reliability assessment method based on performance test data, characterized in that: include: Collect performance test data of various components in rail transit products at different intervals, and process and organize the performance test data to form degradation data; Based on the degradation data, the degradation trajectory method is used to establish the performance degradation model of each component; The model parameter fitting method is used to estimate the parameters of the established performance degradation model and obtain the degradation trajectory model parameters of each sample; According to the degradation trajectory model parameters of each sample and the pre-set failure threshold, the pseudo failure life data of each component is obtained; Conduct life distribution model hypothesis test on pseudo failure life data to determine the appropriate life distribution model; Calculate the reliability index of each component according to the life distribution model; The reliability index of each component is input into the pre-built system reliability model, and the reliability index of the entire rail transit product or high-level components is calculated to complete a comprehensive evaluation of product reliability.

2. The rail transit product reliability assessment method based on performance test data according to claim 1 is characterized in that: The method also includes: before collecting performance test data of each component in the rail transit product at different equal intervals, determining the rail transit product configuration, wherein the rail transit product configuration can manage and display the name of each component in the rail transit product and related information, the subordinate relationship between the components, and the installation location of each component, and perform statistical analysis according to different dimensions included in the configuration.

3. The rail transit product reliability assessment method based on performance test data according to claim 1 is characterized in that: When the degradation trajectory method is applied to establish the performance degradation model of each component, the degradation trajectory function used includes a linear function when the second-order derivative is zero, a concave function when the second-order derivative is positive, and a convex function when the second-order derivative is negative.

4. The rail transit product reliability assessment method based on performance test data according to claim 3 is characterized in that: The linear function is fitted using the following modified linear models: ； ； ； In the formula x i It is the performance degradation index of the product. i For the i Test samples, t For the detection time, α i , β i is a model parameter whose value is estimated through measured performance test data.

5. The rail transit product reliability assessment method based on performance test data according to claim 1 is characterized in that: When determining the failure threshold, the selected degradation indicators include at least one or more main technical indicators. When one or more of them exceed the specified value, it is determined that the rail transit product has failed due to degradation.

6. The rail transit product reliability assessment method based on performance test data according to claim 1 is characterized in that: The model parameter fitting method includes a loss function method or a least squares method.

7. The rail transit product reliability assessment method based on performance test data according to claim 1 is characterized in that: The reliability indicators of each component include reliability, failure probability and failure rate.

8. The rail transit product reliability assessment method based on performance test data according to claim 1 is characterized in that: The life distribution model includes a Weibull distribution model, an exponential distribution model and a normal distribution model.

9. The rail transit product reliability assessment method based on performance test data according to claim 1 is characterized in that: The system reliability model includes a reliability block diagram model, a fault tree analysis model and a Monte Carlo simulation model.

10. The rail transit product reliability assessment method based on performance test data according to claim 1 is characterized in that: When collecting performance test data, the sample size shall be at least 5, and the time interval for recording data shall cover the life cycle of the component, and the corresponding number of recording time points shall be at least 3.

Images