A reliability prediction method and system in a dynamic environment

Abstract

The invention discloses a reliability prediction method in a dynamic environment, including obtaining a reliability index in a static environment by using GJB299C-2006 or MIL-HDBK-217 method; carrying out a Markov process simulation to obtain a track of the Markov process; carrying out a Markov process parameter simulation to obtain a transfer rate matrix of the Markov process; carrying out a dynamic environment modeling and simulation based on the Markov process to obtain a product failure time; repeatedly using a dynamic environment modeling and simulation algorithm based on the Markov process for several times, and obtaining a product failure rate, that is, an MTBF value, according to the obtained result. The invention takes into account the influence of the dynamic environment of the product on the reliability, can improve the existing reliability prediction method that cannot fully reflect the actual situation of the environment where the product is located, obtain a more accurate reliability prediction result, and will play a positive role in improving the product reliability design analysis.

Description

Technical Field

[0001] The present invention belongs to the technical field of reliability, and in particular to a reliability prediction method and system in a dynamic environment. Background Art

[0002] With the development of science and technology and the improvement of human living standards, people's attention to products has shifted from basic functional performance to appearance, reliability and other application experience aspects, and reliability as an important aspect of product quality has received more and more attention. Reliability prediction, as an important means of product reliability design analysis, plays a very positive role in identifying weak links in reliability design and quantitatively evaluating product reliability, and has been widely used in the design stage of military and civilian products.

[0003] The most widely used reliability prediction method is generally to use standards such as MIL-HDBK-217 or GJB299C to carry out predictions. The external environment is one of the main reasons affecting product reliability. This method often assumes that the product is in a static environment when predicting. In fact, the working environment of the product often changes with time and space. Using a static environment to predict product reliability cannot fully reflect the actual situation of the product's environment, and its reliability will have a large deviation. The reliability prediction in a dynamic environment takes into account the changes in the working environment over time and space, reflects the actual situation of the product's environment, and has a positive role in improving the accuracy of reliability prediction. Summary of the invention

[0004] The purpose of the present invention is to provide a reliability prediction method and system in a dynamic environment in order to solve the problems existing in the prior art.

[0005] The technical solution to achieve the purpose of the present invention is: a reliability prediction method in a dynamic environment, the method comprising the following steps:

[0006] Step 1, using GJB299C-2006 or MIL-HDBK-217 method to obtain reliability indicators in a static environment;

[0007] Step 2, conduct a Markov process simulation to obtain a track of the Markov process;

[0008] Step 3, simulate the parameters of the Markov process to obtain the transition rate matrix of the Markov process;

[0009] Step 4: Conduct dynamic environment modeling and simulation based on Markov process to obtain the failure time of a product;

[0010] Step 5, repeatedly use the dynamic environment modeling and simulation algorithm based on the Markov process for several times, and obtain the failure rate of the product, that is, the MTBF value, according to the obtained results.

[0011] A reliability prediction system in a dynamic environment, the system comprising:

[0012] An index acquisition module is used to obtain reliability indexes in a static environment using GJB299C-2006 or MIL-HDBK-217 methods;

[0013] The first simulation module is used to simulate the Markov process and obtain a track of the Markov process;

[0014] The second simulation module is used to simulate the parameters of the Markov process and obtain the transition rate matrix of the Markov process;

[0015] Modeling and simulation module, used to carry out dynamic environment modeling and simulation based on Markov process to obtain the failure time of a product;

[0016] The failure rate acquisition module is used to repeatedly use the dynamic environment modeling and simulation algorithm based on the Markov process for several times, and obtain the failure rate of the product, that is, the MTBF value, according to the obtained results.

[0017] A computer device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the computer program:

[0018] Step 1, using GJB299C-2006 or MIL-HDBK-217 method to obtain reliability indicators in a static environment;

[0019] Step 2, conduct a Markov process simulation to obtain a track of the Markov process;

[0020] Step 3, simulate the parameters of the Markov process to obtain the transition rate matrix of the Markov process;

[0021] Step 4: Conduct dynamic environment modeling and simulation based on Markov process to obtain the failure time of a product;

[0022] Step 5, repeatedly use the dynamic environment modeling and simulation algorithm based on the Markov process for several times, and obtain the failure rate of the product, that is, the MTBF value, according to the obtained results.

[0023] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the following steps:

[0024] Step 1, using GJB299C-2006 or MIL-HDBK-217 method to obtain reliability indicators in a static environment;

[0025] Step 2, conduct a Markov process simulation to obtain a track of the Markov process;

[0026] Step 3, simulate the parameters of the Markov process to obtain the transition rate matrix of the Markov process;

[0027] Step 4: Conduct dynamic environment modeling and simulation based on Markov process to obtain the failure time of a product;

[0028] Step 5, repeatedly use the dynamic environment modeling and simulation algorithm based on the Markov process for several times, and obtain the failure rate of the product, that is, the MTBF value, according to the obtained results.

[0029] Compared with the prior art, the present invention has the following significant advantages: it takes into account the impact of the dynamic environment of the product on reliability, applies Markov theory to describe changes in the external environment, and can correct the reliability indicators in a static environment to obtain more accurate reliability prediction results.

[0030] The present invention is further described in detail below in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 Schematic diagram of sample paths in a dynamic environment.

[0032] Figure 2 This is the simulation algorithm flow chart. DETAILED DESCRIPTION

[0033] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0034] It should be noted that if the embodiments of the present invention involve directional indications (such as up, down, left, right, front, back, etc.), the directional indications are only used to explain the relative position relationship, movement status, etc. between the components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indication will also change accordingly.

[0035] In addition, if there are descriptions involving "first", "second", etc. in the embodiments of the present invention, the descriptions of "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or suggesting their relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, the technical solutions between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in the field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0036] In one embodiment, a reliability prediction method in a dynamic environment is provided, the method comprising the following steps:

[0037] Step 1: Use GJB299C-2006 (or other standards such as MIL-HDBK-217) to obtain reliability indicators in a static environment, such as two reliability indicators or more indicators λ for a good ship environment and a normal ship environment. i ;

[0038] Step 2, conduct a Markov process simulation to obtain a track of the Markov process;

[0039] Step 3, simulate the parameters of the Markov process to obtain the transition rate matrix of the Markov process;

[0040] Step 4: Conduct dynamic environment modeling and simulation based on Markov process to obtain the failure time of a product; Figure 1 is a possible sample path for the system (K=4).

[0041] Step 5, Combine Figure 2 The algorithm gives a simulation process of system failure time. The dynamic environment modeling and simulation algorithm based on Markov process is used repeatedly for several times, such as 10,000 times, 20,000 times or 50,000 times (MATLAB and other tools can be used). The failure rate of the product, that is, the MTBF value, is obtained according to the results.

[0042] Further, in one embodiment, the simulation of the Markov process in step 2 to obtain a track of the Markov process specifically includes:

[0043] For a Markov process X = {X(t), t≥0}, when X(t n )=i n When staying at i n The time of the state {θ n ,n≥1} conditionally independent, and θ n The obedience parameters are The exponential distribution of the state transition probability n≥0, i=1,2,...,K;

[0044] The initial distribution is {P i (0), i∈S} means that the system is in state i at time 0, and the transition rate matrix Q={q ij The simulation of the Markov chain X={X(t),t≥0} orbit can be divided into the following three parts: First, the initial distribution P i (0); then simulate the residence time {θ n ,n≥1}; finally, the state transfer Conduct simulations;

[0045] The specific simulation steps are as follows:

[0046] Step 1: Take two independent sequences of uniformly distributed random variables {U n ,n≥0} and {V n ,n≥1};

[0047] Step 2: If U0 satisfies:

[0048]

[0049] Then take X(0)=i0 as the initial state;

[0050] Step 3: Take out U1 and make as the time X(t) stays at X(0) = i0;

[0051] Step 4: Take out V1, if i1∈S makes V1 satisfy:

[0052]

[0053] Then take X(t1)=X(θ1)=i1;

[0054] Step 5: Take out U2 and make As the dwell time of X(t) in the state X(t1)=i1, take t2=θ1+θ2 as the second jump time;

[0055] Step 6: Take out V2, if i2∈S makes V2 satisfy:

[0056]

[0057] Then take X(t2)=i2;

[0058] ...Continue with the above steps

[0059] Step 7: Remove the Un ,make As X(t) stays at X(t n-1 )=i n-1 The residence time of the state, then take

[0060] Step 8: Remove V n , if i n ∈S makes V n satisfy

[0061]

[0062] Then take X(t n )=i n ;

[0063] By repeating the above steps, you can simulate a track of X.

[0064] Furthermore, in one embodiment, the simulation of Markov process parameters is carried out in step 3 to obtain the transition rate matrix of the Markov process, and the specific process includes:

[0065] For the transfer rate matrix Q = {q ij}, assuming q ij =q ij (β), which depends on b independent parameters; β1,β2,...,β b , that is, β=(β1,β2,...,β b ); Obviously, the transition probability matrix P(t) of the Markov process {X(t), t≥0} = {p ij (t)} is a complex function of these parameters;

[0066] For statistical inference, including statistical inference of parameters, the method depends on the form of the data, that is, different data forms will lead to different statistical inference methods of parameters. First, assume that the data form of the statistical inference used is as follows:

[0067] Data assumptions: Assume that there are several performance-degraded devices, i.e. individuals, observed at the following times:

[0068] t0<t1<t2<…<t n

[0069] Here, the so-called “being observed” refers to the observation time t i , the discrete state of the degraded equipment can be observed;

[0070] Let n ijl Indicates that at time t l-1 Observe that the system is in state i and at time t lThe number of individuals or devices that observe the system in state j is simply expressed as:

[0071] n ijl =#{X(t l )=j,X(t l-1 )=i},i,j=1,2,...,k.

[0072] Where l = 0, 1, ..., n;

[0073] Therefore, based on time t i The likelihood function of each individual, i.e., the state of the device, about the parameter β is given as:

[0074]

[0075] Because of the homogeneity of time, assuming that w l =t l -t l-1 , l=1,2,...,n, then:

[0076]

[0077] Here, the whole process will be implicitly ij (w l β in β, that is, p ij (w l ; β) = p ij (w l ), so the maximum likelihood estimate of β is It can be given by maximizing its log-likelihood function, namely:

[0078] or

[0079] Use the quasi-Newton algorithm to maximize the above log-likelihood function. The specific process of the algorithm is as follows:

[0080] Step 1: Order u=1,2,...,b;

[0081] Step 2: Let vector S(θ) = (S1(θ), S2(θ), ..., S b (θ) T ;

[0082] Step 3: Order (#{Z(t l-1 )=i});

[0083] Step 4: Order u,v=1,2,...,b;

[0084] Step 5: Let the matrix M(θ) = (M uv (θ) b×b ;

[0085] Step 6: Select the initial value θ0 = (θ 1,0 ,θ 2,0 ,...,θ b,0 );

[0086] Step 7: Perform iterative calculations according to the following iterative formula to the hth step.

[0087] θ t+1 =θ t +M -1 (θ t )S(θ t ),t=0,1,...,h.

[0088] Step 8: When the required accuracy is met, stop the calculation and give the maximum likelihood estimate of θ That is, ||θ h -θ h-1 ||<ε, ||·|| is a specified norm, and ε>0 is a specified precision.

[0089] Furthermore, in one embodiment, the step 4 of carrying out the dynamic environment modeling and simulation based on the Markov process specifically includes:

[0090] (1) Assume that the system has K operating environments, denoted by environments 1, 2, ..., K. The transition rules between environments are described by a time-homogeneous continuous-time Markov process {X(t), t ≥ 0}. Let S = {1, 2, ..., K} represent the state space of the process, and Q = {q ij} is the transfer rate matrix, i = 1, 2, ..., K, j = 1, 2, ..., K;

[0091] (2) In environment i, the failure time of the system is subject to parameter λ i The exponential distribution of i That is, the failure rate of the system in environment i, i = 1, 2, ..., K.

[0092] In one embodiment, a reliability prediction system in a dynamic environment is provided, the system comprising:

[0093] An index acquisition module is used to obtain reliability indexes in a static environment using GJB299C-2006 or MIL-HDBK-217 methods;

[0094] The first simulation module is used to simulate the Markov process and obtain a track of the Markov process;

[0095] The second simulation module is used to simulate the parameters of the Markov process and obtain the transition rate matrix of the Markov process;

[0096] Modeling and simulation module, used to carry out dynamic environment modeling and simulation based on Markov process to obtain the failure time of a product;

[0097] The failure rate acquisition module is used to repeatedly use the dynamic environment modeling and simulation algorithm based on the Markov process for several times, and obtain the failure rate of the product, that is, the MTBF value, according to the obtained results.

[0098] For the specific limitations of the reliability prediction system in a dynamic environment, please refer to the limitations of the reliability prediction method in a dynamic environment mentioned above, which will not be repeated here. Each module in the above-mentioned reliability prediction system in a dynamic environment can be implemented in whole or in part through software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0099] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the computer program:

[0100] Step 1, using GJB299C-2006 or MIL-HDBK-217 method to obtain reliability indicators in a static environment;

[0101] Step 2, conduct a Markov process simulation to obtain a track of the Markov process;

[0102] Step 3, simulate the parameters of the Markov process to obtain the transition rate matrix of the Markov process;

[0103] Step 4: Conduct dynamic environment modeling and simulation based on Markov process to obtain the failure time of a product;

[0104] Step 5, repeatedly use the dynamic environment modeling and simulation algorithm based on the Markov process for several times, and obtain the failure rate of the product, that is, the MTBF value, according to the obtained results.

[0105] For the specific limitations of each step, please refer to the limitations of the reliability prediction method in a dynamic environment mentioned above, which will not be repeated here.

[0106] In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0107] Step 1, using GJB299C-2006 or MIL-HDBK-217 method to obtain reliability indicators in a static environment;

[0108] Step 2, conduct a Markov process simulation to obtain a track of the Markov process;

[0109] Step 3, simulate the parameters of the Markov process to obtain the transition rate matrix of the Markov process;

[0110] Step 4: Conduct dynamic environment modeling and simulation based on Markov process to obtain the failure time of a product;

[0111] Step 5, repeatedly use the dynamic environment modeling and simulation algorithm based on the Markov process for several times, and obtain the failure rate of the product, that is, the MTBF value, according to the obtained results.

[0112] For the specific limitations of each step, please refer to the limitations of the reliability prediction method in a dynamic environment mentioned above, which will not be repeated here.

[0113] In summary, the present invention takes into account the impact of the product's dynamic environment on reliability, can improve the existing reliability prediction method that cannot fully reflect the actual situation of the product's environment, obtain more accurate reliability prediction results, and will play a positive role in improving product reliability design analysis.

[0114] The above shows and describes the basic principles, main features and advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A reliability prediction method in a dynamic environment, characterized in that: The method comprises the following steps: Step 1, using GJB299C-2006 or MIL-HDBK-217 method to obtain reliability indicators in a static environment; Step 2, conduct a Markov process simulation to obtain a track of the Markov process; Step 3, simulate the parameters of the Markov process to obtain the transition rate matrix of the Markov process; Step 4: Conduct dynamic environment modeling and simulation based on Markov process to obtain the failure time of a product; Step 5, repeatedly use the dynamic environment modeling and simulation algorithm based on the Markov process for several times, and obtain the failure rate of the product, that is, the MTBF value, according to the obtained results; The simulation of the Markov process described in step 2 is carried out to obtain a track of the Markov process, which specifically includes: For a Markov process X = {X(t), t≥0}, when X(t n )=i n When staying at i n The time of the state {θ n ,n≥1} conditionally independent, and θ n The obedience parameters are The exponential distribution of the state transition probability The initial distribution is {P i (0), i∈S} means that the system is in state i at time 0, and the transition rate matrix Q={q ij The simulation of the Markov chain X={X(t),t≥0} orbit can be divided into the following three parts: First, the initial distribution P i (0); then simulate the residence time {θ n ,n≥1}; finally, the state transfer Conduct simulations; The specific simulation steps are as follows: Step 1: Take two independent sequences of uniformly distributed random variables {U n ,n≥0} and {V n ,n≥1}; Step 2: If U0 satisfies: Then take X(0)=i0 as the initial state; Step 3: Take out U1 and make as the time X(t) stays at X(0) = i0; Step 4: Take out V1, if i1∈S makes V1 satisfy: Then take X(t1)=X(θ1)=i1; Step 5: Take out U2 and make As the dwell time of X(t) in the state X(t1)=i1, take t2=θ1+θ2 as the second jump time; Step 6: Take out V2, if i2∈S makes V2 satisfy: Then take X(t2)=i2; ...Continue with the above steps Step 7: Remove the U n ,make As X(t) stays at X(t n-1 )=i n-1 The residence time of the state, then take Step 8: Remove V n , if i n ∈S makes V n satisfy Then take X(t n )=i n ; By repeating the above steps, you can simulate a track of X.

2. The reliability prediction method in a dynamic environment according to claim 1, characterized in that: Step 3 is to simulate the Markov process parameters and obtain the transition rate matrix of the Markov process. The specific process includes: For the transfer rate matrix Q = {q ij }, assuming q ij =q ij (β), which depends on b independent parameters; β1,β2,...,β b , that is, β=(β1,β2,...,β b ); Obviously, the transition probability matrix P(t) of the Markov process {X(t), t≥0} = {p ij (t)} is a complex function of these parameters; For statistical inference, including statistical inference of parameters, the method depends on the form of the data, that is, different data forms will lead to different statistical inference methods of parameters. First, assume that the data form of the statistical inference used is as follows: Data assumptions: Assume that there are several performance-degraded devices, i.e. individuals, observed at the following times: t0<t1<t2<…<t n Here, the so-called "observed" refers to the observation time t i , the discrete state of the degraded equipment can be observed; Let n ijl Indicates that at time t l-1 Observe that the system is in state i and at time t l The number of individuals or devices that observe the system in state j is simply expressed as: n ijl =#{X(t l )=j,X(t l-1 )=i},i,j=1,2,...,k, Where l = 0, 1, ..., n; Therefore, based on time t i The likelihood function of each individual, i.e., the state of the device, about the parameter β is given as: Because of the homogeneity of time, assuming that w l =t l -t l-1 , l=1,2,...,n, then: Here, the whole process will be implicitly ij (w l β in β, that is, p ij (w l ; β) = p ij (w l ), so the maximum likelihood estimate of β is It can be given by maximizing its log-likelihood function, namely: or The quasi-Newton algorithm is used to maximize the above log-likelihood function. The specific process of the quasi-Newton algorithm is as follows: Step 1: Order Step 2: Let vector S(θ) = (S1(θ), S2(θ), …, S b (θ) T ; Step 3: Order (#{Z(t l-1 )=i}); Step 4: Order Step 5: Let the matrix M(θ) = (M uv (θ) b×b ; Step 6: Select the initial value θ0 = (θ 1,0 ,θ 2,0 ,...,θ b,0 ); Step 7: Perform iterative calculations according to the following iterative formula to the hth step. i t+1 =θ t +M -1 (i t )S(θ t ),t=0,1,...,h, Step 8: When the required accuracy is met, stop the calculation and give the maximum likelihood estimate of β That is, ||β h -β h-1 ||<ε, ||·|| is a specified norm, and ε>0 is a specified precision.

3. The reliability prediction method in a dynamic environment according to claim 1, characterized in that: Step 4 describes the Markov process-based dynamic environment modeling and simulation, which specifically includes: (1) Assume that the system has K operating environments, denoted by environments 1, 2, ..., K. The transition rules between environments are described by a time-homogeneous continuous-time Markov process {X(t), t ≥ 0}. Let S = {1, 2, ..., K} represent the state space of the process, and Q = {q ij } is the transfer rate matrix, i = 1, 2, ..., K, j = 1, 2, ..., K; (2) In environment i, the failure time of the system is subject to parameter λ i The exponential distribution of i That is, the failure rate of the system in environment i, i = 1, 2, ..., K.

4. A reliability prediction system in a dynamic environment based on the method according to any one of claims 1 to 3, characterized in that: The system comprises: An index acquisition module is used to obtain reliability indexes in a static environment using GJB299C-2006 or MIL-HDBK-217 methods; The first simulation module is used to simulate the Markov process and obtain a track of the Markov process; The second simulation module is used to simulate the parameters of the Markov process and obtain the transition rate matrix of the Markov process; Modeling and simulation module, used to carry out dynamic environment modeling and simulation based on Markov process to obtain the failure time of a product; The failure rate acquisition module is used to repeatedly use the dynamic environment modeling and simulation algorithm based on the Markov process for several times, and obtain the failure rate of the product, that is, the MTBF value, according to the obtained results.

5. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 3 are implemented.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 3 are implemented.

Images