Accelerated life test method for reliability analysis of vehicle-mounted control equipment

Abstract

According to the acceleration test method for reliability analysis of the vehicle-mounted control equipment, an acceleration model suitable for the vehicle-mounted control equipment is established according to a determined characteristic stress set; determining a stress value of an acceleration test and a stress value under an actual application environment, and calculating an acceleration factor; determining acceleration test time and test criteria according to an expected reliability target of the vehicle-mounted control equipment and the calculated acceleration factor; and executing the test and recording data in the test process, and performing qualitative evaluation and quantitative analysis on the reliability level of the vehicle-mounted control equipment based on the test data. According to the acceleration test method provided by the invention, the reliability analysis of the rail transit vehicle-mounted control equipment can be accurately and efficiently completed within a short test time, and the problems of accuracy and economy encountered when the high-reliability and long-service-life vehicle-mounted control equipment is subjected to reliability test evaluation are solved.

Description

Technical Field

[0001] The invention relates to the technical field of accelerated testing, in particular to an accelerated testing method for reliability analysis of rail transit vehicle-mounted control equipment. Background Art

[0002] Accelerated reliability testing is an important branch in the field of reliability engineering. It aims to shorten the test time by increasing the stress level of product tests or increasing the frequency of alternating stress application, and to evaluate, analyze and improve the reliability of products with limited samples, time and usage costs. This technology is based on the theory of accelerated damage. It equates the damage caused to the product by the stress value in the test with the cumulative damage caused by the expected stress and time during the product life cycle. It uses the mathematical relationship related to the physical failure law, namely the acceleration model, to convert the reliability information obtained in an accelerated environment that exceeds the normal stress level, and obtain an estimated value of the reliability characteristics of the product under the actual stress level. In short, accelerated testing is to shorten the test time by increasing the test stress or stress frequency while keeping the failure mechanism unchanged, and infer the reliability level of the product in actual application through data analysis.

[0003] Since it was proposed in the 1960s, reliability accelerated testing technology has developed vigorously and has gradually been applied in many fields such as automotive electronics, consumer electronics, and communication equipment. It has formed a variety of accelerated test models and test methods, such as the Arrhenius model and the Coffin-Manson model, which have become important methods for reliability analysis and evaluation of electronic equipment. Due to the high complexity, wide application scenarios and high reliability requirements of rail transit on-board control equipment, there are the following difficulties in conducting accelerated tests: 1) Accelerated stress lacks specificity: Accelerated stress depends on the understanding of the failure mechanism of the product. Due to the high complexity of the product and the many types of failure mechanisms in the equipment, it is difficult to extract effective stress loads. Improper selection may result in the failure of the main failure mechanism in the actual application environment to be effectively accelerated, while some failure modes that are not easy to occur are accelerated, affecting the accuracy of the analysis and the efficiency of the test; 2) Inaccurate accelerated stress value: The premise of the accelerated test is that the failure mechanism under accelerated stress is consistent with the failure mechanism under normal stress. Excessive accelerated stress may change the failure mechanism of the product and affect the accuracy of the analysis. Excessive stress may lead to insufficient cumulative damage to the product and affect the efficiency of the test; 3) The accelerated test time is long: Rail transit on-board control equipment usually adopts a 2x2 take 2 redundant structure, which has a high mission reliability. Limited by the number of test samples, even if the accelerated test method is used to verify the reliability index, the test time is relatively long, affecting the test efficiency and the economy of the test. Summary of the invention

[0004] The present invention provides an accelerated life test method for reliability analysis of rail transit vehicle-mounted control equipment, aiming to solve the accuracy and economy problems encountered in performing accelerated reliability tests on high-reliability and long-life vehicle-mounted control equipment.

[0005] The present invention provides an accelerated life test method for reliability analysis of vehicle-mounted control equipment, and the specific steps are as follows:

[0006] 1) According to the determined characteristic stress set, a fault physical model that can reflect the relationship between the fault occurrence and the characteristic stress set is selected from the mature fault physical models, and an acceleration model suitable for vehicle control equipment is established;

[0007] 2) Determine the stress value of the accelerated test and the stress value under the actual application environment, substitute them into the acceleration model formula, and calculate the acceleration factor;

[0008] 3) Determine the accelerated test time and test criteria based on the expected reliability target of the vehicle control equipment and the calculated acceleration factor;

[0009] 4) Execute the test and record the data during the test, and make a qualitative evaluation and quantitative analysis on the reliability level of the vehicle-mounted control device based on the test data.

[0010] The technical advantages of the scheme of the present invention are as follows: In order to enhance the pertinence of the accelerated test stress, the present invention proposes to use the historical failure information of the equipment to determine the key reliability components under the actual operating conditions of the equipment, and to identify the key stresses that play a major role in their failure by analyzing the failure mechanism of the key reliability components. At the same time, these key stresses are screened in combination with the mission profile in the product life cycle, and the stresses that exist under actual working conditions and play a major role in failure are extracted as test acceleration stresses to improve the accuracy of the test.

[0011] In order to improve the accuracy of the acceleration stress value, the present invention proposes to use the step test theory to find the appropriate acceleration stress value through experiments, and to increase the acceleration stress value as much as possible without exceeding the test environment capability and without changing the product failure mechanism, thereby shortening the time and improving the test efficiency.

[0012] Aiming at the high reliability requirements of vehicle-mounted control equipment and combining the characteristics of redundant structure, the present invention uses Markov state transition diagram to convert the task reliability index of the redundant system into the MTBF of a single set of equipment. Under the premise that the acceleration factor remains unchanged, the test time can be greatly shortened, the test cost can be reduced, and the test economy can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 is a step diagram of the accelerated test method of the present invention;

[0014] Figure 2 Identify steps for reliability critical components;

[0015] Figure 3 This is a schematic diagram of TCR historical fault data processing;

[0016] Figure 4 It is the Pareto chart of TCR historical failure data;

[0017] Figure 5 Extraction diagram for characteristic stress set;

[0018] Figure 6 Determine the step test diagram for TCR temperature stress value;

[0019] Figure 7 It is the TCR Markov state transition diagram. DETAILED DESCRIPTION

[0020] The present invention is described in detail below in conjunction with specific embodiments. The following embodiments will help technicians further understand the present invention, and the embodiments are only illustrative and should not be construed as any form of limitation of the present invention. It should be pointed out that any modification and improvement made by technicians without departing from the concept of the present invention belongs to the protection scope of the present invention.

[0021] Figure 1 This is a step diagram of the accelerated test method of the present invention, which is applicable to the accelerated test of reliability analysis of vehicle-mounted control equipment. The specific steps are as follows:

[0022] 1) According to the determined characteristic stress set, select a fault physics model that can reflect the relationship between the fault occurrence and the characteristic stress set from the mature fault physics model, and establish an acceleration model suitable for vehicle control equipment. The fault physics model is a mathematical function model that quantitatively reflects the relationship between the fault occurrence and stress, etc., based on the basic physical, chemical, electrical formulas and (or) experimental regression formulas for a specific fault mechanism in fault physics.

[0023] 2) Determine the stress value of the accelerated test and the stress value under the actual application environment, substitute them into the acceleration model formula, and calculate the acceleration factor. For example, if the selected fault physical model is the Arrhenius model reflecting temperature stress, the established acceleration model is as follows:

[0024]

[0025] Where, Ts is the thermodynamic value of the test temperature (i.e., the stress value of the accelerated test), Tu is the thermodynamic value of the actual application environment temperature (i.e., the stress value under the actual application environment), and A is a constant. Once the values ​​of Ts and Tu are determined, the acceleration factor AF can be calculated by substituting them into the above formula;

[0026] 3) According to the expected reliability target of the vehicle control equipment and the calculated acceleration factor, determine the accelerated test time and test criteria. The so-called test criteria means that the test ends when the test time specified in the test plan is reached. When the number of failures is less than or equal to the number of failures specified in the test plan, an acceptance judgment is made, and it is considered that the reliability index of the test sample meets the expected reliability requirements. Otherwise, a rejection judgment is made.

[0027] 4) Carry out the test and record the data during the test, and make qualitative evaluation and quantitative analysis on the reliability level of rail transit on-board control equipment based on the test data.

[0028] When determining the experimental characteristic stress set, not only the stress load that can play a major role in the failure process of key reliability components is considered, but also the mission profile of the product life cycle is combined to screen out the stress that actually exists under actual working conditions. That is, the characteristic stress set is determined by taking the intersection of the characteristic stress set obtained from the failure mechanism research of key reliability components (devices) and the characteristic stress set obtained from the mission profile research.

[0029] When identifying key reliability components, the equipment historical failure information recorded by the FRACAS (Failure Reporting, Analysis, and Corrective Action System) system is used to identify key reliability components. By performing a Pareto chart analysis on the failure data of the equipment in the actual application environment, the failure components whose Pareto chart cumulative curve is close to or exceeds 80% are identified as key reliability components. This can concentrate most of the causes of failures on a few key components, narrow the scope of research on failure mechanisms, and enhance the targeted nature of accelerated stress.

[0030] When taking the test stress value, the step test theory is used to take the test stress intensity value when the working characteristics of the test sample no longer meet the requirements of the technical conditions, but the test sample can still restore the normal working characteristics after the test stress is reduced as the accelerated test stress value. The stress value in the application environment is obtained through mission profile research.

[0031] When determining the accelerated test time, the Markov state transition diagram of the redundant system is constructed to convert the expected reliability index of the redundant system into a single set of MTBF indicators, thereby greatly shortening the test time under the premise of keeping the acceleration factor unchanged.

[0032] Figure 2This is a schematic diagram of the present invention applied to the identification of key reliability components of the track circuit reader (TCR). The FRACAS system collects fault information of the TCR equipment in actual use in the past year, covering routine maintenance fault information, field application fault information and fault information during the test process, including basic information such as the time of the fault, the location of the fault, the type of fault, the location of the fault, the cause of the fault, etc. In order to ensure the comprehensiveness of the data, the FRACAS system recorded all 52 faults that occurred in the TCR in the past year. When using this fault information to identify key reliability components, it is first processed to eliminate invalid data that is not related to the equipment reliability analysis.

[0033] Figure 3 This is a schematic diagram of data processing of TCR historical fault information. Among all 52 faults that occurred in TCR, there were 3 non-related faults caused by improper operation or excessive external stress. Among the remaining 49 related faults, 40 faults were reset and successfully located, and there were 9 invalid fault information, including 6 faults that were not located, 2 faults that were recorded repeatedly, and 1 fault with incorrect information recording. After removing the above non-related faults and invalid fault information, there are 40 valid fault information that can be used to identify reliability-critical components.

[0034] Figure 4 This is a Pareto analysis chart of 40 valid faults recorded by TCR. The pre-processed fault records are classified according to the type of faulty devices, the number of each type of devices is counted and sorted in order from large to small, the proportion of each type of device in the total number of faults and the cumulative failure probability are calculated, and the Pareto chart is drawn. The left vertical axis in the figure represents the number of device failures, the right vertical axis represents the cumulative failure probability of each type of faulty device, and the horizontal axis represents the type of faulty device. The middle part draws a bar chart of each type of device from left to right in the order of the number of faulty devices from large to small, and the upper part of the middle part draws the cumulative failure probability curve in the order of the cumulative failure probability from small to large. According to the 80 / 20 rule, in most cases, most problems (80%) are often caused by a few key factors (20%). This principle is used when identifying key reliability components, that is, a few devices that cause the cumulative failure probability of the equipment to approach or exceed 80% are regarded as key components affecting product reliability. Figure 4 In the Pareto chart, the cumulative failure probability corresponding to the cumulative curve "□" is 82.5%. At this time, the corresponding devices are CPU modules, A / D chips, power chips and communication chips. That is, these four types of devices cause more than 80% of the field failures of TCR equipment. Therefore, these four types of devices are determined to be the "key" devices that affect the reliability of TCR.

[0035] Figure 5Represents the process of extracting accelerated characteristic stress sets in the present invention. For TCR equipment, by analyzing the failure mechanism of four types of reliability key components, namely CPU module, A / D chip, power chip and communication chip, the main stress loads that cause the failure of each type of device can be identified, and all stress loads that cause the failure of the four types of devices are taken as a complete set to obtain a characteristic stress set A that causes the failure of key reliability components. Through the study of the mission profile of the TCR product life cycle, the environmental stress and working stress that actually exist under actual working conditions are extracted as the characteristic stress set B. The intersection of set A and set B is taken to obtain a characteristic stress set C that exists in the actual application environment and acts on the failure mechanism of key reliability components. C is the characteristic stress set that should be considered in the final TCR accelerated test, that is, C = A∩B.

[0036] The characteristic stress set A is obtained by failure mode, mechanism and effect analysis (FMMEA). FMMEA was proposed by Professor Michael Pecht of the University of Maryland in the United States in the 1990s based on the research and application of FMEA technology. It is a failure physics method based on the evaluation of product failure mechanisms. Through FMMEA, the failure mechanism analysis of the key components of TCR reliability, CPU module, A / D chip, power chip and communication chip (as shown in Table 1) was carried out. It can be concluded that the potential failure points of the key components of TCR reliability are chips, adhesive layers, substrates, wire bonds, flip-chip solder joints and shells. The corresponding failure mechanisms are corrosion, electromigration, thermal fatigue, creep, mechanical fatigue and dendrite growth, etc. The main characteristic stress set A extracted is {steady-state temperature, relative humidity, temperature cycle, temperature shock, mechanical vibration, voltage / current, etc.}. Through data collection and on-site investigation, the mission profile of TCR during its life cycle is studied, including the working environment and mission events at different stages such as startup, shutdown, storage, and transportation, and the final characteristic stress set B = {steady-state temperature, relative humidity, mechanical vibration, electromagnetic environment} is extracted.

[0037] Table 1

[0038]

[0039]

[0040] According to the above analysis, the intersection of the two stress sets is taken to obtain the final characteristic stress set of the TCR accelerated test, that is, C = A ∩ B = {steady-state temperature, relative humidity, temperature cycle, temperature shock, mechanical vibration, voltage / current, etc.} ∩ {steady-state temperature, relative humidity, mechanical vibration, electromagnetic environment} = {steady-state temperature, relative humidity, mechanical vibration}. Considering that the coupling between various stresses is relatively complex, it is difficult to simply use a multiplication relationship to simultaneously link them with the life of the product. Therefore, in order to simplify the problem and improve the accuracy of the test, combined with mature fault physics models and engineering implementation experience, the steady-state temperature and relative humidity are finally selected from the set C as the test stress of the TCR accelerated test, and the acceleration model is selected as the Hallberg-Peck model. The acceleration factor formula is expressed as:

[0041]

[0042] Figure 6 This is a schematic diagram of determining the value of accelerated temperature stress of the TCR host through a temperature step test. According to the operating temperature range in the design data and existing experience, the device failure inside the TCR device will not occur when the ambient temperature is 45°C. In order to shorten the test time and improve efficiency, the initial test temperature starts at 45°C, and is maintained for 2 hours. After the TCR temperature stabilizes, it is maintained for 10 minutes, and then the functional performance test is performed. If no abnormality is found after the test, the step temperature is increased to 55°C and the above steps are repeated. When the ambient temperature rises to 75°C, in order to more accurately obtain the working limit temperature of the product, the temperature rise step is changed to 5°C and the above steps are repeated. When the ambient temperature rises to 90°C and is maintained for 2 hours, the internal ambient temperature of the host reaches about 100°C. At this time, the test finds that the device function is abnormally reset. When the temperature drops to 85°C, the internal ambient temperature of the test is about 94°C, and the test equipment can start working normally. It is judged that a value between 85°C and 90°C is the working limit temperature of the TCR host. In order to ensure the smooth progress of the accelerated test, the final temperature stress value is 85°C for accelerated test. Similarly, the step test can be used to determine the humidity stress value. Combined with the mission profile study and test results, the humidity stress is finally taken as 85% relative humidity. According to the determined acceleration factor model and the stress value of the actual environment, the acceleration factor is calculated to be 1040.4.

[0043] According to TB / T 3533-2018 Track Circuit Reader (TCR), the MTBSF of the TCR host should not be less than 1×10 6h. According to the timed truncation test scheme in TB / T 2468-2018 "Evaluation Method for Reliability Requirements of Railway Signal Products", if the number of test samples is 3, the manufacturer's risk α and the user's risk β are both 10%, and the acceleration factor is 1040, then the accelerated test time needs to be at least 1038.5 hours. When the number of failures that affect the operation service of the three devices during the 1038.5-hour accelerated test is less than or equal to 2, an acceptance judgment is made, that is, the MTBSF of the TCR host meets 1×10 6 h is reliable, otherwise a rejection decision is made.

[0044] In order to further shorten the test time and reduce the test cost, the reliability target of the TCR host is converted into the MTBF value requirement of a single set. Figure 7 is the Markov state transition diagram of the redundant system, where state 2 means both sets are working properly, state 1 means one set is working properly, state 0 means both sets cannot work properly, λ is the failure rate of a single set, λ C is the common cause failure rate of the two sets, μ is the maintenance rate of a single set, then the calculation formula for the host MTBSF is:

[0045]

[0046] The TCR maintenance time is generally no more than 48 hours. However, considering the uncertainties such as the arrival time at the site and the availability of spare parts, the maintenance time is conservatively taken as 1000 hours, that is, μ = 0.001 / h. Assuming that the MTBF of a single system is 1×10 5 h, that is, λ = 1 × 10 -5 / h, the TCR common cause failure rate is evaluated according to the common cause failure analysis method provided in GB / T 20438.6-2017 "Functional safety of electrical / electronic / programmable electronic safety-related systems Part 6", and the common cause failure factor scoring result β is taken as 5%, then λ C =β×λ=5×10 -7 / h, substituting into the above formula, we get TCR host MTBSF = 1.44 × 10 6 h, meets industry standard 1×10 6 h. Therefore, the accelerated test can be carried out on a single system according to MTBF = 1 × 10 5 h. Under the premise that other test conditions remain unchanged, the test time can be shortened to about 103.8 h. The corresponding test criterion is that when the number of failures of a single set of equipment during the 103.8-hour accelerated test of the three equipment is less than or equal to 2, an acceptance judgment is made, that is, the MTBF of a single TCR set meets 1×10 5 h is reliable, otherwise a rejection decision is made.

[0047] According to the above test conditions and test time, the accelerated test was organized and implemented. During the test, the test samples were tested and the fault data were collected. Finally, the test results were judged and the test data were analyzed. During the accelerated test, the three TCR devices did not fail within 103.8h, that is, the number of failures was less than 2, so an acceptance judgment was made, that is, the MTBF of a single TCR device met 1×10 5 h reliability requirements, the TCR host MTBSF meets the industry standard 1×10 6 h's requirements.

[0048] The test data is analyzed by statistical principles, including point estimation and confidence interval estimation of reliability characteristic values:

[0049] The single set MTBF point estimate is calculated as follows:

[0050]

[0051] Where T is the cumulative test time, T = 3 × 103.8 × 1040 = 3.24 × 10 5 h;

[0052] C is the confidence level, which is 0.9. Substituting it into the above formula, the point estimate of MTBF for a single set is 1.41×10 5 h, and substitute it into the host MTBSF calculation formula, and we get the MTBSF of the TCR host = 1.81 × 10 6 h.

[0053] The one-sided interval estimate of a single set of MTBF is calculated as follows:

[0054]

[0055] Where T is the cumulative test time, 2T = 2 × 3 × 103.8 × 1040 = 6.48 × 10 5 h, is the upper (right) quantile, C is the confidence level, which is also 0.9, r is the number of faults, and in this case r is 0. Substituting it into the above formula, we get: Then the lower limit of the one-sided interval estimation of a single set of MTBF = 1.41×10 5 h, substitute into the host MTBSF calculation formula, MTBSF = 1.81 × 10 6 h.

[0056] The beneficial effects of the present invention are:

[0057] The accelerated test method proposed in the present invention can accurately and efficiently complete the reliability analysis of rail transit vehicle-mounted control equipment in a shorter test time, and solve the accuracy and economy problems encountered in the reliability test and evaluation of high-reliability and long-life vehicle-mounted control equipment.

[0058] Improve accuracy: By performing Pareto chart analysis on historical failure data in real environments, it is possible to more accurately identify key components that affect the reliability of equipment in actual applications. Combined with mission profile research within the product life cycle and failure mechanism analysis of key reliability components, characteristic stresses that exist in actual use environments and play a major role in the failure mechanism of key reliability components can be extracted as accelerated test stresses, thereby improving test accuracy.

[0059] Improve efficiency: Determine the value of accelerated stress according to the step test theory, combine the acceleration model and task profile, and increase the value of accelerated stress as much as possible without changing the failure mechanism, shorten the test time, and improve the test efficiency. At the same time, use Pareto chart analysis to concentrate most of the causes of failures on a few key components, which can reduce the scope of failure analysis and the types of accelerated test stress.

[0060] Reduce costs: Based on the structural characteristics of rail transit vehicle control equipment, which is mostly dual-set redundant, the reliability indicators of redundant equipment are converted into single-set reliability requirements through Markov state transition diagrams. Under the premise that the acceleration factor does not change, the test time can be significantly reduced and the test cost can be reduced.

[0061] Improve reliability and safety: By improving the accuracy of the test, more accurate test data can be obtained, and then a more accurate qualitative evaluation and quantitative analysis of the reliability level of the vehicle control equipment can be made. The reliability problems found in the test can be used as reliability improvement information and introduced into the subsequent design and production links to continuously improve the reliability and safety of vehicle control equipment.

[0062] The above is only a preferred embodiment of the present invention and is not intended to limit the protection scope of the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An accelerated life test method for reliability analysis of vehicle-mounted control equipment, the specific steps of which are as follows: 1) According to the determined characteristic stress set, a fault physical model that can reflect the relationship between the fault occurrence and the characteristic stress set is selected from the mature fault physical models, and an acceleration model suitable for vehicle control equipment is established; 2) Determine the stress value of the accelerated test and the stress value under the actual application environment, substitute them into the acceleration model formula, and calculate the acceleration factor; 3) Determine the accelerated test time and test criteria based on the expected reliability target of the vehicle control equipment and the calculated acceleration factor; 4) Execute the test and record the data during the test, and make a qualitative evaluation and quantitative analysis on the reliability level of the vehicle-mounted control device based on the test data.

2. The accelerated life test method according to claim 1, characterized in that: Step 1) When determining the characteristic stress set of the accelerated test, not only the stress load that can play a major role in the failure process of the key reliability components is considered, but also the mission profile of the product life cycle is combined to screen out the stress that actually exists under the actual working conditions, that is, the characteristic stress set of the accelerated test is determined by taking the intersection of the characteristic stress set obtained from the failure mechanism research of the key reliability components and the characteristic stress set obtained from the mission profile research.

3. The accelerated life test method according to claim 2, characterized in that: When identifying the reliability-critical components, the equipment historical failure information recorded by the FRACAS system is used to identify the reliability-critical components. By performing a Pareto chart analysis on the failure data of the equipment in an actual application environment, the failure components whose Pareto chart cumulative curves are close to or exceed 80% are identified as the reliability-critical components.

4. The accelerated life test method according to claim 1, characterized in that: Step 2) When determining the accelerated test stress value, use the step test theory to take the test stress intensity value when the working characteristics of the test sample no longer meet the requirements of the technical conditions but the test sample can still restore the normal working characteristics after the test stress is reduced as the accelerated test stress value.

5. The accelerated life test method according to claim 1, characterized in that: Step 3) When determining the accelerated test time, the expected reliability index of the redundant system is converted into a single set of MTBF index by constructing a Markov state transition diagram of the redundant system, thereby shortening the accelerated test time under the premise of keeping the acceleration factor unchanged.

6. The accelerated life test method according to claim 1, characterized in that: Step 4) During the qualitative evaluation and quantitative analysis, the reliability index of the redundant system is evaluated and analyzed by recording and analyzing the test data of a single set of equipment and combining it with the reliability index formula of the redundant system.

Images