A vehicle reliability assessment method that balances the inspection risks of both parties

Abstract

The present invention proposes a vehicle reliability assessment method that balances the risks of both parties, solves the problem that the existing frequency school reliability assessment method is insufficiently applied in engineering, conducts detailed research on its theoretical system and ideas, sorts out the assessment ideas of the frequency school, and conducts analysis and research on currently commonly used mathematical software and function libraries. The present invention and method can avoid complicated theoretical foundations and provide a new way for designers to conduct vehicle reliability assessment.

Description

Technical Field

[0001] The invention belongs to the field of vehicle reliability assessment work, and in particular relates to a vehicle reliability assessment method that weighs the inspection risks of both parties. Background Art

[0002] Vehicle reliability assessment has always been one of the important work items of military product reliability work. This work item is based on probability statistical analysis technology. By analyzing the failure rate data of components, parts and equipment related to the system composition and various test data, the system reliability is integrated layer by layer to check whether the products at all levels have reached the specified reliability goals. It plays a vital role in system design trade-off decisions and test plan selection. The vehicle reliability assessment method mainly includes two parts: 1. Development of reliability assessment plan; 2. Evaluation of reliability test data.

[0003] The formulation of reliability assessment plans is currently the most widely used in engineering based on the frequency school of probability theory. The frequency school believes that for a batch of samples, its distribution F(X,θ) is certain, that is, θ is certain, but θ is unknown. Although the theoretical system of the frequency school is complete, there are several problems in its application: (1) The theoretical system of the frequency school based on probability theory is complex and too abstract, difficult to understand, and brings great difficulties to product designers to effectively and quickly formulate reasonable assessment methods; (2) The formulation of reliability assessment plans often requires comprehensive consideration of the risks of the manufacturer and the user. However, there is currently no practical optimization method in the field of vehicle reliability testing that weighs the inspection risks of both parties; (3) After the test, it is impossible to accurately and quickly assess the actual reliability level of the vehicle. Summary of the invention

[0004] The technical problem solved by the present invention is: in order to solve the defects that the methods in the prior art are too abstract and difficult to quickly formulate evaluation methods, and there is no practical optimization method that weighs the inspection risks of both parties and cannot evaluate the actual reliability level of the vehicle, the present invention proposes a vehicle reliability evaluation method that weighs the inspection risks of both parties.

[0005] The technical solution of the present invention is: a vehicle reliability assessment method for weighing the risks of both parties, comprising the following steps:

[0006] Step 1: Use the MATLAB function model to solve the test mileage multiplication factor K;

[0007] Step 2: Use the MATLAB function model to calculate the production risk α and the user risk β, and determine the reliability assessment plan;

[0008] Step 3: Build a fast calculation function for MMBF and solve the minimum acceptable value of the vehicle's mean mileage between failures (MMBF) based on actual test data.

[0009] A further technical solution of the present invention is: the step 1 comprises the following sub-steps:

[0010] Step 1.1: Collect the quantitative requirements of vehicle reliability;

[0011] Step 1.2: Establish the relationship model of the test mileage multiplier factor K, that is,

[0012] Step 1.3: Build a fast calculation model for the test mileage multiplication factor K, and select MATLAB "chi2inv" as the calculation function of K, that is: K = chi2inv (γ, 2*f+2) / 2;

[0013] Step 1.4: Preliminarily calculate the test mileage multiplication factor K in the selected interval, and calculate the test mileage multiplication factor K for different values ​​of the allowable number of failures f;

[0014] Step 1.5: After completing an interval operation in 1.4, proceed to step 2 to calculate the risks of both parties. If the actual risk is basically the same as the required risk value, terminate this process; otherwise, return to step 1.4 to continue searching for the next space.

[0015] A further technical solution of the present invention is: the f is set to 0-20 as a step gradient, and a total of four intervals are taken.

[0016] A further technical solution of the present invention is: the step 2 comprises the following sub-steps:

[0017] Step 2.1: Based on the reliability information collected in 1.1, use the model D = L U / L D Calculate the test discrimination factor D as one of the input parameters of step 2;

[0018] Step 2.2: Obtain the other two input parameters, multiplication factor K and allowable fault number f, through step 1;

[0019] Step 2.3: Construct a simplified statistical risk mathematical model for vehicles:

[0020]

[0021] Step 2.4: Construct a basic model for fast calculation of vehicle simplified statistical risk:

[0022]

[0023] Step 2.5: Calculate the risk of a certain interval (f, K) in step 1 through step 2.4, and check it with the required producer risk α and user risk β, and select the group (f, K) that is closest to the risks of both parties;

[0024] Step 2.6: Use the user risk value β' obtained in step 2.5 as the input for further adjusting the confidence level, that is, the actual confidence level γ' = 1-β';

[0025] Step 2.7: Based on the latest γ', call step 1.4 again to obtain the latest test mileage multiplier factor K';

[0026] Step 2.8: Call step 2.5 again according to the latest test mileage multiplier factor K' to calculate the final actual risks α” and β” of both parties;

[0027] Step 2.9: Select the latest confidence γ', risk α", β", and according to the test mileage multiplier factor K', the best reliability assessment scheme for this vehicle is obtained as (L, f).

[0028] A further technical solution of the present invention is: the step 3 comprises the following sub-steps:

[0029] Step 3.1: According to the evaluation scheme (L, f) obtained in step 2, conduct verification tests and collect the actual total mileage L' of the same model vehicle on land. When L'=L, if the actual number of failures f'≤f, the vehicle reliability passes the test, otherwise it fails the test; the minimum acceptable value of the actual mean failure interval mileage MMBF is L' D Calculated by step 3.2; if the actual test mileage L'≠L, then directly calculate the actual MMBF L' through step 3.2 D , determine whether the requirements are met;

[0030] Step 4.2: Similar to step 1.3, select MATLAB "chi2inv" as L' D The fast operation function is: L' D =2*L / chi2inv(γ,2*f+2) / 2, calculate the reliability level of the actual vehicle according to the model, and compare it with the required value to determine whether the current reliability level of the vehicle meets the requirements;

[0031] Step 4.3: If the requirements are not met, the test may be continued as appropriate, or the verification test may be repeated after the fault is corrected.

[0032] Effects of the Invention

[0033] The technical effect of the present invention is that the present invention constructs a rapid assessment and optimization process that weighs the inspection risks of both parties, forming an effective relationship chain of various mathematical models. The assessment process and the applied mathematical model are continuously optimized and adjusted through the relevant data provided by the national standard and the historical data of previous models, and finally a more accurate and clear assessment method is formed. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 : Flowchart of Method A

[0035] Figure 2 :Method B Flowchart

[0036] Figure 3 :Method C Flowchart DETAILED DESCRIPTION

[0037] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise" and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as limiting the present invention.

[0038] See also Figure 1-Figure 3 ,This scheme describes a vehicle reliability assessment method that balances the inspection risks of both parties. This method is mainly composed of three sub-methods A, B, and C collaboratively designed.

[0039] Method A: The method for calculating the test mileage multiplier factor K based on the MATLAB function model. The method flow is shown in Figure 1 , mainly including the following steps:

[0040] Step A.1: Collect the quantitative requirements for vehicle reliability. For the vehicle field, reliability parameters include basic reliability parameters and mission reliability parameters. For example, basic reliability parameters generally include the minimum acceptable value L of the mean mileage between failures (MMBF) D , specified value L U , one-sided confidence γ, allowable number of failures f (to be determined), producer risk α, user risk β;

[0041] Step A.2: Establish a relationship model for the test mileage multiplier factor K. According to the actual working characteristics of the vehicle, determine the test scheme using exponential distribution mileage, and the randomized optimal confidence upper limit λ of the failure rate uand twice the mileage L follow the chi-square distribution, that is And the exponential distribution of L D =1 / λ u ,but Definition K = L / L D ,Right now

[0042] Step A.3: Build a fast calculation model for the test mileage multiplication factor K, and select MATLAB "chi2inv" as the calculation function of K, that is: K = chi2inv (γ, 2*f+2) / 2;

[0043] Step A.4: Preliminary calculation of the test mileage multiplication factor K in the selected interval. Calculate the test mileage multiplication factor K for different values ​​of the allowable failure number f. According to engineering experience, it is usually sufficient to take f from 0 to 20. In actual operation, the step gradient is 5, that is, 0 to 5, 6 to 10, 11 to 15, and 16 to 20, a total of four intervals;

[0044] Step A.5: After completing an interval operation in A.4 each time, enter method B to calculate the risks of both parties. If the actual risk is basically the same as the required risk value, terminate this process. Otherwise, enter step A.4 to continue searching for the next space.

[0045] Method B: A method based on MATLAB function model to quickly calculate the production risk α and the user risk β, and determine the reliability assessment scheme. The method flow is shown in Figure 2 , including the following steps:

[0046] Step B.1: Based on the reliability information collected in A.1, the model D = L U / L D Calculate the test discrimination factor D as one of the input parameters of method B;

[0047] Step B.2: Obtain the other two input parameters, the multiplication factor K and the allowable fault number f, through method A (these two parameters should be calculated through method A and correspond one to one);

[0048] Step B.3: Construct a simplified mathematical model of vehicle statistical risk. Based on the statistical test model of the frequency school exponential distribution model, that is: and D=L U / L D , K=L / L D , then the statistical test model of vehicle reliability assessment is simplified to:

[0049] Step B.4: Construct a basic model for fast calculation of vehicle simplified statistical risk, select the MATLAB function "factorial" as the calculation model, and then the mathematical model in step B.3 is simplified to the basic calculation model:

[0050] Step B.5: Calculate the risk of a certain interval (f, K) of method A through step B.4, and check it with the required producer risk α and user risk β, and select the group (f, K) that is closest to the risks of both parties;

[0051] Step B.6: Use the user risk value β' obtained in step B.5 as input for further adjusting the confidence level, i.e., the actual confidence level γ' = 1-β';

[0052] Step B.7: Based on the latest γ', call step A.4 again to obtain the latest test mileage multiplier factor K';

[0053] Step B.8: Call step B.5 again according to the latest test mileage multiplier factor K' to calculate the final actual risks α" and β" of both parties;

[0054] Step B.9: Select the latest confidence γ', risk α", β", and obtain the best reliability assessment scheme for this vehicle (L, f) based on the test mileage multiplier factor K'.

[0055] Method C: A method for calculating the minimum acceptable value of the vehicle mean mileage between failures (MMBF) based on actual test data. The method flow is shown in Figure 3 , including the following steps:

[0056] Step C.1: According to the evaluation scheme (L, f) obtained by method B, conduct verification tests and collect the actual total mileage L' of the same model vehicle on land. When L'=L, if the actual number of failures f'≤f, the vehicle reliability passes the test; otherwise, it fails the test and the minimum acceptable value of the actual mean failure interval mileage MMBF is L' D Calculated through step C.2; if the actual test mileage L'≠L, then directly calculate the actual MMBF L' through step C.2 D , determine whether the requirements are met;

[0057] Step C.2: Similar to step A.3, select MATLAB "chi2inv" as L' D The fast operation function is: L' D =2*L / chi2inv(γ,2*f+2) / 2, calculate the reliability level of the actual vehicle according to the model, and compare it with the required value to determine whether the current reliability level of the vehicle meets the requirements;

[0058] Step C.3: If the requirements are not met, the test may be continued as appropriate, or the verification test may be repeated after the fault is corrected. DETAILED DESCRIPTION

[0060] Step 1: Collect the quantitative reliability requirements of the vehicle. For example, the reliability requirements of a certain type of vehicle are: the minimum acceptable value of the mean mileage between failures (MMBF) is 3000km, the specified value is 6000km, the one-sided confidence level γ is 0.8, the manufacturer's risk α is 0.2, and the user's risk β is 0.2. Figure 1 The technical process of method A shown in the figure first calculates the test mileage multiplication factor K for f=0 to 5. After calculation, it can be obtained that the test mileage multiplication factor K corresponding to different numbers of faults is (0, 1.6094), (1, 2.9943), (2, 4.279), (3, 5.515), (4, 6.721), (5, 7.906).

[0061] Step 2: Execution Figure 2 The technical process of method B shown in the figure is that D=2 is calculated from the information in step 1, and steps B.1 to B.6 in the technical solution of "method B" are executed to calculate the risks corresponding to different (f, K). The number of failures closest to the risks of both parties is (f, K)=(5, 7.906), and its risk value is: the production party risk α=0.207568, β=0.199999;

[0062] Step 3: Execute steps B.7 to B.10 to obtain the final reliability assessment plan (L, f). For example, the reliability assessment plan for this time is (24041km, 5).

[0063] Step 4: Carry out the test according to the evaluation scheme obtained in step 3, and evaluate its reliability level according to the actual test data obtained during the field test. If a certain type of vehicle carries out reliability tests on two vehicles at the same time, each vehicle actually runs 25,000 kilometers, the total test mileage is 50,000 kilometers, and the two vehicles have a total of 7 related failures, then the reliability level of this type of vehicle is determined by "Method C" in the technical solution. The minimum acceptable value of the mean mileage between failures (MMBF) is 4886km>3000km, which meets the requirements.

Claims

1. A vehicle reliability assessment method that balances the risks of both parties, characterized in that: The following steps are involved: Step 1: Use the MATLAB function model to solve the test mileage multiplication factor K; in It includes the following sub-steps: Step 1.1: Collect the quantitative requirements of vehicle reliability; Step 1.2: Establish the relationship model of the test mileage multiplier factor K, that is, Where L is the mileage, L D is the minimum acceptable value of the mean mileage between failures; Step 1.3: Build a fast calculation model for the test mileage multiplication factor K, and select MATLAB "chi2inv" as the calculation function of K, that is: K = chi2inv (γ, 2*f+2) / 2; where f is the allowable number of failures and γ is the one-sided confidence level; Step 1.4: Preliminarily calculate the test mileage multiplication factor K in the selected interval, and calculate the test mileage multiplication factor K for different values ​​of the allowable number of failures f; Step 1.5: After completing an interval operation in 1.4, proceed to step 2 to calculate the risks of both parties. If the actual risk is roughly equivalent to the required risk value, terminate this process. Otherwise, return to step 1.4 to continue searching for the next space. Step 2: Use the MATLAB function model to calculate the production risk α and the user risk β, and determine the reliability assessment plan; It includes the following sub-steps: Step 2.1: Based on the reliability information collected in 1.1, use the model D = L U / L D Calculate the test discrimination factor D as one of the input parameters of step 2; where L U It is the specified value of the mean mileage between failures; Step 2.2: Obtain the other two input parameters, multiplication factor K and allowable fault number f, through step 1; Step 2.3: Construct a simplified statistical risk mathematical model for vehicles: Step 2.4: Construct a basic model for fast calculation of vehicle simplified statistical risk: Step 2.5: Calculate the risk of a certain interval (f, K) in step 1 through step 2.4, and check it with the required producer risk α and user risk β, and select the group (f, K) that is closest to the risks of both parties; Step 2.6: Use the user risk value β' obtained in step 2.5 as the input for further adjusting the confidence level, i.e., the actual confidence level γ' = 1-β'; Step 2.7: Based on the latest γ', call step 1.4 again to obtain the latest test mileage multiplier factor K'; Step 2.8: Call step 2.5 again according to the latest test mileage multiplier factor K' to calculate the final actual risks α” and β” of both parties; Step 2.9: Select the latest confidence γ', risk α", β", and obtain the best reliability assessment scheme for this vehicle (L, f) based on the test mileage multiplier factor K'; Step 3: Build a fast calculation function of MMBF and solve the minimum acceptable value of the vehicle's mean mileage between failures MMBF based on actual test data; It includes the following sub-steps: Step 3.1: According to the evaluation scheme (L, f) obtained in step 2, conduct verification tests and collect the actual total mileage L' of the same model vehicle on land. When L'=L, if the actual number of failures f'≤f, the vehicle reliability passes the test, otherwise it fails the test; the minimum acceptable value of the actual mean failure interval mileage MMBF is L' D Calculated by step 3.2; if the actual test mileage L'≠L, then directly calculate the actual MMBF L' through step 3.2 D , determine whether the requirements are met; Step 4.2: Similar to step 1.3, select MATLAB "chi2inv" as L' D The fast operation function is: L' D =2*L / chi2inv(γ,2*f+2) / 2, calculate the reliability level of the actual vehicle according to the model, and compare it with the required value to determine whether the current reliability level of the vehicle meets the requirements; Step 4.3: If the requirements are not met, the test may be continued as appropriate, or the verification test may be repeated after the fault is corrected.

2. A vehicle reliability assessment method for weighing the risks of both parties as claimed in claim 1, characterized in that: The f ranges from 0 to 20, with a step gradient of 5, and there are four intervals of 0 to 5, 6 to 10, 11 to 15, and 16 to 20.

Images