A method and system for predicting product concept design reliability

Abstract

The application discloses a kind of product concept design reliability prediction methods, comprising: S1 is carried out to target product structure classification, obtains the key components of target product and the design life value of target product;S2 according to the key components in S1 is for the target product Virtual test matter element model R is built, and the virtual test matter element model R includes the key components of the target product theoretical design parameter;S3 is based on the virtual test matter element model R in S2, carries out iterative analysis and obtains multiple groups about target product Accumulative damage value and corresponding virtual life value;S4 according to the virtual life value corresponding when the first occurrence cumulative damage value is greater than or equal to 1 in S3 as comparison value, compared with the design life of target product, to judge the reliability of target product.The application also provides a kind of system suitable for the prediction method.The system is simple to operate, and it is convenient for user to carry out quick judgment to the reliability of product concept design scheme.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of product design software, and in particular to a product concept design scheme reliability prediction method and system. BACKGROUND

[0002] At the initial stage of new product development, the demand side does not have clear requirements for reliability, and often only focuses on development speed, function implementation and user-friendly interaction, etc. When put into use, a large number of reliability problems are found, increasing the difficulty of later maintenance, and even leading to a series of safety problems. In fact, product concept design has a great influence on the reliability of later mass production, and about 70% of product failures are hidden in the concept design stage.

[0003] However, few people pay attention to this aspect, the main reasons are:

[0004] (1) The product in the concept design stage lacks a relatively mature digital model and physical prototype. For many enterprise R&D departments, only part of the demand parameters and structure parameters, it is difficult to predict the reliability of the product.

[0005] Many high-tech enterprises, such as Tesla, Waymo, and Baidu, have begun to simplify the digital model of the initial product, and use computer programs to simulate actual working conditions to complete virtual testing, but for small and medium-sized enterprises, it is difficult to complete the digital prototype and perform a large number of test cases in a short time.

[0006] (2) Enterprises are reluctant to provide internal digital models to third-party companies, so it is difficult to predict the reliability of the concept model using existing commercial software (such as ANSYS, nCode, etc.).

[0007] At present, there are two ways to complete virtual reliability prediction, one is to build a life model based on a failure physical model, and the other is a data-driven reliability evaluation method.

[0008] The academic literature "Fuzzy Reliability Prediction and Scheme Evaluation of Conceptual Design Products" [J]. Mechanical Science and Technology, 2005 (08): 981-984. discloses a fuzzy reliability prediction for a conceptual design product, quantifies the possibility of bottom event occurrence of a fault tree in fuzzy language description as an element of a fuzzy set, and uses physical programming to evaluate the conceptual design scheme, thereby providing a new way for conceptual design scheme evaluation, but this scheme only defines the initial fuzzy interval, and does not consider the influence of design parameter iteration evolution on system reliability.

[0009] Patent document CN113094805A discloses a structure analysis parameterized model generation system and a generation method thereof, which comprises a formulating module for formulating parameterized modeling specifications, an implementation module for researching key technologies for parameterized modeling implementation, and a development module for developing parameterized modeling software.

[0010] Currently, the reliability prediction in the conceptual design stage is based on complete digital models, and the weakest position reliability is predicted by combining the simulation results and the life curve of the product structure. However, the conceptual design stage often does not have a basic digital model, and it is impossible to carry out reliability evaluation.

[0011] Cai Wen, a researcher, proposed and established the discipline of artificial intelligence, and the extension correlation function is an important part of extension.

[0012] The extension correlation function can make it possible to describe the quantitative change and qualitative change by establishing the correlation function on the real axis, which is essentially different from the characteristic function of Cantor set and the membership function of Fuzzy set.

[0013] Common correlation functions include simple correlation functions, elementary correlation functions, interval correlation functions and discrete correlation functions. The use of extension correlation function theory can effectively solve the contradiction that the existing set theory cannot handle the same thing with different things. SUMMARY

[0014] To solve the above problems, the present application provides a prediction method for the reliability of product conceptual design scheme, which has fast operation speed, does not require specific product design parameters and is suitable for various industries.

[0015] A prediction method for the reliability of product conceptual design, comprising:

[0016] S1 classifies the target product by structure, obtains the key components of the target product and the design life value of the target product;

[0017] S2 constructs a virtual test matter-element model R for the target product according to the key components in S1, wherein the virtual test matter-element model R contains the theoretical design parameters of the key components of the target product; Wherein M represents the target product, C i represents the function of the target product, X iwhere i represents the key components affecting the function, and i is the total number of key components.

[0018] For example, an electric vehicle is classified into power supply, power and control, where the battery affects the power supply, the electric drive affects the power output, and the controller affects the vehicle control. Based on the above classification, a virtual test matter-element model R of the electric vehicle is constructed.

[0019] Virtual test matter-element model of electric vehicle

[0020] S3. Based on the virtual test matter-element model R in S2, iterative analysis is performed to obtain a plurality of sets of accumulated damage values and corresponding virtual life values of the target product.

[0021] S4. The virtual life value corresponding to the first accumulated damage value greater than or equal to 1 in S3 is taken as a comparison value, which is compared with the design life of the target product to determine the reliability of the target product.

[0022] Preferably, the construction method of the virtual test matter-element model R in S2 is as follows:

[0023] S2.1. Obtain the key components of the target product, and set a corresponding virtual test interval (a, b) for each key component, where a is the minimum theoretical design parameter value and b is the maximum theoretical design parameter value.

[0024] S2.2. Build a virtual test matter-element model R of the key components X i S2.3. Perform secondary classification, including the performance related to the key components, the design parameters related to the components, and the corresponding virtual test intervals:

[0025] Test matter-element model of key components

[0026] where m i represents the performance related to the key components, c j represents the theoretical design parameters affecting the performance, x j represents the virtual test interval of the theoretical design parameters, and j is the total number of theoretical design parameters.

[0027] For example, the battery of an electric vehicle: the battery performance mainly includes charging efficiency and power decay,

[0028]

[0029] Preferably, the iterative analysis in S3 is based on the calculation analysis of the virtual test points of the key components, the virtual test points are randomly selected in a range greater than zero, when the virtual test points are randomly selected outside the virtual test interval, the damage value of the product under low load or overload working condition is simulated and calculated; when the virtual test points are randomly selected within the virtual test interval, the damage value of the product under the theoretical working condition is simulated and calculated, both cases are recorded for calculating the virtual life value of the product under the comprehensive working condition.

[0030] Preferably, the specific process of the iterative analysis is as follows:

[0031] S3.1 randomly selects a group of virtual test points of the key components of the target product, and calculates the corresponding correlation degree of each virtual test point according to the extension correlation function theory

[0032]

[0033] Wherein x mj is the virtual test point of the key component.

[0034] Finally, the cumulative correlation degree P j corresponding to all virtual test points of the group of key components is calculated and obtained

[0035] S3.2 based on the cumulative correlation degree P j obtained in S3.1, a determination is made:

[0036] When 0≤P j ≤1, continue to repeat S3.1, and record the iteration number;

[0037] When 1<P j , the damage value = 1 / P j is calculated, and the damage value and the iteration number are recorded.

[0038] S3.3 based on the damage value obtained in S3.2 and the corresponding iteration number, the cumulative damage value is calculated, that is, the damage values corresponding to the iteration numbers are sequentially accumulated according to the size of the iteration numbers;

[0039] At the same time, the iteration number of this time is recorded as the virtual life value corresponding to the cumulative damage value.

[0040] Preferably, when the cumulative damage value is less than 1, the cumulative damage value and the corresponding virtual life value are recorded, and the next round of iterative analysis is continued; otherwise, the iterative analysis is terminated.

[0041] Preferably, the method for judging the reliability of the target product in S4 is as follows: when the comparison value is greater than the design life value of the target product, the scheme is qualified; when the comparison value is less than or equal to the design life value of the target product, the scheme is unqualified.

[0042] A system for predicting reliability of product concept design scheme, comprising:

[0043] A virtual test point module for providing random virtual test points for key components;

[0044] A virtual test interval module for converting theoretical design parameters of key components into virtual test intervals;

[0045] A cumulative correlation degree module for iterative analysis of key components;

[0046] An accumulated damage calculation module for accumulated damage value of target product and determination of whether to continue iterative analysis;

[0047] A reliability evaluation module for comparison between results of iterative analysis and virtual use frequency value of target product, and output of conclusion based on comparison results.

[0048] Specific use process: technical grading is performed on a target product, key components are obtained, theoretical design parameters of each key component are converted into corresponding virtual test intervals through the virtual test interval module; iterative analysis is started, a virtual test point is provided for each key component through the virtual test point module; then, one round of iterative analysis is performed on the target product through the cumulative correlation degree module; accumulated damage value and corresponding virtual life value are obtained through the accumulated damage calculation module, and when the accumulated damage value appears for the first time and is greater than or equal to 1, the iterative analysis of this round is terminated; through the reliability evaluation module, the virtual life value corresponding to the first time when the accumulated damage value is greater than or equal to 1 in this round is selected as a comparison value, and comparison is performed between the comparison value and the service life of the target product to obtain the iterative analysis result of this round; further, three rounds of iterative analysis are required and comparison is performed with the service life of the target product in sequence, and when the results of the three rounds are all qualified, a conclusion that the scheme is qualified is output; if the result of one time is unqualified, a conclusion that the scheme is unqualified is output, and a designer is prompted to perform debugging.

[0049] Compared with the prior art, the present application has the beneficial effects that: the present application adopts extension modeling and correlation function theory, avoids the difficulty of non-uniform dimensions under multiple evaluation characteristics, attributes multiple evaluation indexes to a unique objective function, and greatly improves the calculation speed; further, the system provided by the present application is compatible with the current mainstream design software, and makes up for the deficiency that some software cannot perform scheme reliability prediction. BRIEF DESCRIPTION OF DRAWINGS

[0050] Figure 1 A system module diagram is provided for the present application;

[0051] Figure 2 A flowchart of the prediction method of the present application. DETAILED DESCRIPTION

[0052] As Figure 1 shown, the present application provides a system for predicting product concept design scheme reliability, which mainly comprises:

[0053] a virtual test point module for providing random virtual test points for key components;

[0054] a virtual test interval module for converting the theoretical design parameters of key components into virtual test intervals;

[0055] a cumulative correlation degree module for iterative analysis of key components;

[0056] an accumulated damage calculation module for accumulated damage value of the target product and determining whether to continue iterative analysis;

[0057] a reliability evaluation module for comparing the results of iterative analysis with the design life value of the target product and outputting conclusions based on the comparison results.

[0058] The above functional modules are compiled into independent executable programs and placed in the main program directory of existing computer-aided design software (such as AutoCAD) to facilitate designers to preliminarily complete product design according to user requirements and quickly judge the reliability of the design scheme.

[0059] Specific use process:

[0060] Classify the structure of the target product to obtain key components, and convert the theoretical design parameters of each key component into corresponding virtual test intervals through the virtual test interval module;

[0061] Start iterative analysis and provide a virtual test point for each key component through the virtual test point module;

[0062] Perform one round of iterative analysis on the target product through the cumulative correlation degree module, and obtain the cumulative damage value and the corresponding virtual life value through the accumulated damage calculation module. When the cumulative damage value first appears greater than or equal to 1, terminate this round of iterative analysis;

[0063] Select the virtual life value corresponding to the first time when the cumulative damage value is greater than or equal to 1 in this round as the comparison value through the reliability evaluation module, and compare it with the design life value of the target product to obtain the iterative analysis result of this round;

[0064] Further, three rounds of iterative analysis are required and are compared with the service life of the target product in turn. When the results of the three rounds are all qualified, the conclusion that the scheme is qualified is output. If the result of one time is unqualified, the conclusion that the scheme is unqualified is output, and the designer is prompted to debug the design parameters.

[0065] As Figure 2 shown, a new energy vehicle development stage concept design scheme is taken as an example to analyze and predict the new energy vehicle:

[0066] S1 is based on the function of new energy vehicles to classify the structure, mainly including power supply, power and control, among which the battery affects the power supply performance, the electric drive affects the power output, and the controller affects the control performance of the vehicle.

[0067] According to the above technical classification, a new energy vehicle virtual test matter element model R is established:

[0068] New energy vehicle virtual test matter element model

[0069] According to the content in the new energy vehicle virtual test matter element model R, a test key component model X is established.

[0070] The main design parameters affecting the performance of the battery include the charging efficiency and the rate of power decay of the battery:

[0071] Test battery model

[0072] The main design parameters affecting the power output include the no-load back electromotive force and insulation of the motor, and the wear resistance of the reducer:

[0073] Test electric drive model

[0074] Test electric drive model X3 = | Reducer performance wear resistance (50, 80) |

[0075] The main design parameters affecting the control performance of the vehicle include the junction temperature of the IGBT chip:

[0076] Test controller model X4 = | IGBT performance junction temperature (120, 125) |

[0077] The design life of the target product is converted into the running frequency value T1, which is 16330 times.

[0078] S2 generates a first group of virtual test points by system randomization:

[0079]

[0080] S3 calculates the correlation degree of each virtual test point x mj and the virtual test interval

[0081]

[0082] According to formula (2), the correlation degree of each virtual test point Cumulative correlation P j :

[0083]

[0084] S4 calculates the correlation degree of each component according to formula (1) and formula (2). Cumulative correlation P with the target product j :

[0085]

[0086] Where N is the number of iterations, and according to the principle of iterative analysis, when P1>1, the damage value D1 needs to be calculated:

[0087]

[0088] And record the damage value D and the iteration number N (0.005266238, 1) of the virtual test points;

[0089] Since the cumulative damage value ∑D1=0+D1<1, the iterative analysis continues, and the cumulative damage value ∑D1 and the virtual lifetime value t1=1 are recorded.

[0090]

[0091] Based on the principle of iterative analysis, when P2 > 1, the damage value D2 needs to be calculated.

[0092]

[0093] And record the damage value D and the iteration number N (0.001756726,2) of the virtual test points;

[0094] Since the cumulative damage value ∑D2=0+D1+D2<1, the iterative analysis continues, and the cumulative damage value ∑D2 and the virtual lifetime value t2=2 times are recorded.

[0095] The iterative analysis continues until the first cumulative damage value ∑D≥1 is calculated, at which point the current iteration is terminated.

[0096] The results of this round of iterative analysis are shown in the table below:

[0097] virtual lifetime value t i ]]> Damage value D i ]]> Cumulative damage value ∑D 1 0.005266238 0.005266238 2 0.001756726 0.007022964 101 0.034220532 0.041243496 3209 0.107964667 0.149208163 6235 0.215662388 0.364870551 8562 0.311672663 0.676543214 12642 0.209081630 0.885624844 13269 0.137373214 1.022998058

[0098] When the virtual lifetime value t i When the cumulative damage value ∑D > 1 at t = 13269, it indicates that the current iteration analysis has ended, and t is taken as... i =13269 is the comparison value.

[0099] According to the comparison value and the design life T1 of the target product, that is, T1 > t i It is indicated that the current round of iterative analysis result is unqualified.

[0100] After repeating the three rounds of iterative analysis according to the above steps, the system determines the three rounds of results. If all the three rounds are qualified, the conclusion that the scheme is qualified is output. If one round is unqualified, the conclusion that the scheme is unqualified is output, and the designer is prompted to debug the design scheme.

Claims

1. A method for predicting the reliability of product concept design, characterized in that, include: S1 performs structural classification of the target product to obtain the key components of the target product and the design life value of the target product. S2 constructs a virtual test matter-element model R for the target product based on the key components in S1. The virtual test matter-element model R includes the theoretical design parameters of the key components of the target product. S3, based on the virtual test matter-element model R in S2, performs iterative analysis to obtain multiple sets of cumulative damage values ​​and corresponding virtual lifetime values ​​for the target product. This iterative analysis is based on virtual test points of key components, which are randomly selected within a range greater than zero. The specific process is as follows: S3.1 Randomly select a set of virtual test points for key components, calculate the correlation degree ρ corresponding to each virtual test point according to the extension correlation function theory, and finally calculate the cumulative correlation degree P corresponding to all virtual test points of the set of key components. j ; S3.2 The cumulative correlation degree P obtained based on S3.1 j Perform a judgment: When 0≤P j If the value is ≤1, continue repeating S3.1 and record the number of iterations; When 1 <P j At that time, calculate the damage value and record the damage value and the number of iterations. S3.3 Based on the damage value and the corresponding iteration number obtained in S3.2, calculate the cumulative damage value, that is, accumulate the corresponding damage value according to the size of the iteration number; At the same time, the number of iterations is recorded as the virtual lifetime value corresponding to the cumulative damage value in that iteration; S4 uses the virtual lifetime value corresponding to the first occurrence of a cumulative damage value greater than or equal to 1 in S3 as a comparison value, and compares it with the design lifetime value of the target product to determine the reliability of the target product.

2. The method for predicting the reliability of product concept design according to claim 1, characterized in that, The specific method for constructing the virtual test matter-element model R in S2 is as follows: S2.1 Obtain the key components of the target product and set a corresponding virtual test interval (a, b) for each key component, where a is the minimum theoretical design parameter value and b is the maximum theoretical design parameter value; S2.2 performs a secondary classification of key components, including the performance of key components, the design parameters related to the components, and the corresponding virtual test range.

3. The method for predicting the reliability of product concept design according to claim 1, characterized in that, The degree of correlation Calculated using extension correlation function theory: Where x mj For key components, virtual test points are defined, where a represents the minimum theoretical design parameter value and b represents the maximum theoretical design parameter value.

4. The method for predicting the reliability of product concept design according to claim 1, characterized in that, When the accumulated damage value is less than 1, record the accumulated damage value and the corresponding virtual lifetime value, and continue to the next round of iterative analysis; otherwise, terminate the iterative analysis.

5. The method for predicting the reliability of product concept design according to claim 1, characterized in that, The specific method for determining the reliability of the target product in S4 is as follows: if the comparison value is greater than the design life value of the target product, the solution is qualified; if the comparison value is less than or equal to the design life value of the target product, the solution is unqualified.

6. A system for predicting the reliability of product concept design schemes, characterized in that, The steps for performing the prediction method for product concept design reliability as described in any one of claims 1 to 5 include: The virtual test point module is used to provide random virtual test points for key components; The virtual test interval module is used to convert the theoretical design parameters of key components into virtual test intervals; The cumulative correlation module is used for iterative analysis of key components; The cumulative damage calculation module is used to accumulate damage values ​​and determine whether iterative analysis should continue. The reliability evaluation module compares the results of iterative analysis with the virtual usage count of the target product and outputs conclusions based on the comparison results.

7. The system for predicting the reliability of product concept design schemes according to claim 6, characterized in that, The reliability evaluation module needs to sequentially judge the results of multiple rounds of iterative analysis. If the conclusions of multiple rounds are all qualified, the solution is displayed as qualified; if one round is unqualified, the solution is displayed as unqualified and the designer is reminded to debug.

Images