CN108170892A - A kind of fault modes and effect analysis method that emulation is deduced based on accident dynamic - Google Patents
A kind of fault modes and effect analysis method that emulation is deduced based on accident dynamic Download PDFInfo
- Publication number
- CN108170892A CN108170892A CN201711248522.3A CN201711248522A CN108170892A CN 108170892 A CN108170892 A CN 108170892A CN 201711248522 A CN201711248522 A CN 201711248522A CN 108170892 A CN108170892 A CN 108170892A
- Authority
- CN
- China
- Prior art keywords
- event
- function
- state
- failure
- time
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
Abstract
The present invention is a kind of fault modes and effect analysis method that emulation is deduced based on accident dynamic, this method is directed to the characteristics of analyzed object, build the function box of each module of minimum indenture level, constructing function block diagram, crash rate is added in function module and to carry out fail-safe analysis different from traditional, the method of the present invention is to each function box, state is included to its internal act structure, the accident dynamic of the elements such as event deduces simulation model, and it is the event addition delay in model, the attributes such as crash rate, the operation mechanism described inside function module true to nature.Then by logic, the time relationship between constructing function box, all function box are connected into an entirety.Finally by the method for discrete event dynamic simulation, the dynamic running process of simulation system finds out the failure event of system and its probability of generation, and does further failure effect and hazard analysis.Solve the problems, such as that conventional failure pattern and impact analysis method can not carry out temporal logic analysis and combination function failure analysis.
Description
Technical field
The present invention is a kind of fault modes and effect analysis method that emulation is deduced based on accident dynamic, belongs to complication system
Reliability analysis technology field.
Background technology
Fault mode, influence and HAZAN fault modes and effect analysis (Failure Mode andEffects
Analysis it is) to sum up to come in engineering practice, based on fault mode, the dividing as target using failure influence or consequence
Analysis technology.The influence that it works to system by analyzing the different faults of each component part one by one, it is comprehensively thin in identification design
Weak link and key project, and the reliability to evaluate and improving system design provides essential information.
Fault modes and effect analysis method is widely adopted in reliability of complex system field at present.Traditional event
Barrier pattern and impact analysis are analyzed, and using single standalone feature as starting point, when analysis considers that more is quiet between each component
State logical relation seldom considers dynamic time restriction relation between each component, only with the technical ability and warp of fail-safe analysis personnel
It tests, is limited by human cognitive ability, it is difficult to which all possible behavior of forecasting system (including normal and abnormal behaviour) is easily dredged
It leaks certain thrashing states or judges the influence of thrashing by accident, current highly complex and integrated system can not be adapted to.
The highly complex system such as current aerospace space flight is real-time system mostly, and the correctness of system is not only in real-time system
Dependent on its logic behaviour, and dependent on its when step response.For example, in a real-time system, an activity performs
It is too late or carry out operation using out of season data, even if this activity or the function completed of operation are correct, but
It is otiose for whole system, there may come a time when it is harmful.In recent years, the air crash of generation is more by each
The failure of temporal logic generation is not met during a component interaction.Ex-post analysis is found, when this kind of failure occurs, individual feature component
There is no problem, but is led to the problem of during multicompartment interaction.
Such issues that solve, foreign countries start to introduce based on the model testing in Formal Verification complicated airborne for 2001
The reliability assessment field of system.Whether the method detection aforementioned system model of adoption status space search meets specific property
Matter, and provide counter-example when system is unsatisfactory for property.But complete formalization method is promoted at home to be had the following problems：1) mould
Completely the Method Modeling based on Formal Verification is complicated and difficulty is larger during type is examined, and can only be used in part system, complete
Full formalization method is difficult to popularize in an all-round way to use at home.2) Model Checking ensures that model is correct, but can not reflect
Go out the failure probability that model applies hardware, the failure probability of real system actual use situation can not be provided.
To solve the above problems, research one kind can be analyzed during multiple component interactions caused by temporal logic mistake
Failure, but the fault modes and effect analysis method that can reduce modeling complexity is very necessary.
Invention content
The present invention be exactly for the above-mentioned prior art situation and designing provide and a kind of emulation deduced based on accident dynamic
Fault modes and effect analysis method, the purpose is to the method by discrete event dynamic simulation, the dynamic operations of simulation system
Process finds out the failure event of system and its probability of generation, and does further failure effect and hazard analysis.Solves biography
The problem of system fault modes and effect analysis method can not carry out temporal logic analysis and combination function failure analysis.
The purpose of the present invention is achieved through the following technical solutions：
This kind deduces the fault modes and effect analysis method of emulation based on accident dynamic, it is characterised in that：This method
Step is as follows：
Step 1: according to product agreement, product demand and the design documentation of analyzed object, analyzed object is described
Composition and function, according to complete machine, system, subsystem, Field Replaceable Unit, internal field replaceable units level will be analyzed pair
As decomposing layer by layer, need to decompose minimum indenture level according to fault modes and effect analysis；
Step 2: for each module of minimum indenture level, a function box is established, constructing function block diagram,
In functional block diagram, a function box represents a function, and function can include subfunction, according to mutual between function box
It relies on and hierarchical relationship, is connected with the line of no arrow between function box, represent the physical relation between each function box, function
The descriptor format of box is：" function function titles
Feature content defines
end function”；
Step 3: for each function box in functional block diagram, its internal act structure accident dynamic is deduced imitative
True mode, the simulation model are a detailed descriptions for including following 10 kinds of elements, including subfunction (C), input variable (I),
State variable (S), output variable (O), original state (R), failure state (F), event (E), state conversion (T), transmission function
(L), event correlation relationship (ER), is reduced to<C,I,S,O,R,F,E,T,L,ER>It is as follows to specifically describe method for ten tuples：
1. C represents subfunction in ten tuples, the function box comprising subfunction is needed to describe this, otherwise need not be retouched
It states.The subfunction that function box includes is write out during description successively.The descriptor format of subfunction is：" function title：Function type ".
Such as：function func1
C
E1:subfunc1；E2:subfunc1；E3:subfunc1；
end function
Wherein func1 contains function E1, E2 and E3 of three subfunc1 types；
2. I represents input variable in ten tuples, each function box is required for describing this.Each function module includes one
A or multiple input variable.Specific form is " name variable:Types of variables:In ", such as：input1:float:in.If become
Measurer has fixed initial value, then types of variables could alternatively be variate-value, such as input2:[0,1]:in；
3. S represents state variable in ten tuples, the box for not including subfunction is needed to describe this, otherwise need not be retouched
It states.State variable refers to some condition or situation in the life cycle of function, and state is that function performs a series of activities
As a result, comprising one or more state variables in each function module, when meeting corresponding conditions, state can be carried out and turned
It changes.The form of state variable is " name variable:[variate-value]:S ", such as：state1:[working,failed,repair]:
s；
4. R represents original state in ten tuples, the box for not including subfunction is needed to describe this, otherwise need not be retouched
It states.Only there are one original state in each functional block diagram, original state R is one kind in S, and R ∈ S are work(when emulation starts
It can state in which.The form of original state is：" Status Name:=state value ", such as state1:=working；
5. the F in ten tuples represents failure state, the box for not including subfunction is needed to describe this, otherwise need not
Description.In each functional block diagram there are one or multiple failure states, failure state be one kind in S, F ∈ S, in simulation process
In, if reaching the state, record is needed into the time of the state and calculates failure probability of happening.The form of failure state
For：" Status Name:=state value ", such as state1:=failed；
6. O represents output variable in ten tuples, each function box is required for describing this.Each function module includes one
A or multiple output variables.The form of output variable is：" name variable:Types of variables:Out ", such as：output1:float:
out；
7. E represents event in ten tuples, the box for not including subfunction is needed to describe this, otherwise without description.
Event can be converted with triggering state, and state conversion must satisfy corresponding switch condition, and when the conditions are met, event triggering just may be used
It is converted with driving condition.Functioning as event increases delay time for state conversion.Event is broadly divided into the invention
Two major class：It is time delay event and temporal event respectively, wherein time delay event is chance event respectively again comprising two classes
(probability-distribution function for carrying parameter) and set time delay event.Temporal event is also to send out at once respectively comprising two classes
It makes trouble part and prerequisite event.The form of chance event in delay event is：" event title (delay:=probability distribution
Function (failure probability lambda)) ", such as：failure1(delay:=exponential (1E-10)).In delay event
Set time delay event form be:" event title (delay:=Dirac (delay time)) ", such as：failure2
(delay:=Dirac (2)).Temporal event need to only specify event title, and the condition of prerequisite event will turn in state
Described in changing, therefore the form of temporal event is：" event title ".Such as：failure3；
8. T represents state conversion in ten tuples, the box for not including subfunction is needed to describe this, otherwise need not be retouched
It states.State conversion is a kind of relationship between two state variables, represents that object will perform certain action in the state of source, and
Enter dbjective state when some particular event occurs or certain specific conditions meet.State conversion form be：" state name
Claim：Source state value [switch condition is true] |-event title->Status Name:=dbjective state ", wherein switch condition is optional, when
When switch condition is empty, event occurs can driving condition conversion.Such as：(state1=working) |-failure->
state1:When=failed represents that failure events occur, the state of state1 is converted to failed by working；
9. L represents transmission function in ten tuples, each function box is required for describing this.Transmission function describes three kinds and patrols
The relationship of collecting：It is respectively：
The logical relation between output and input and state variable inside 9.1 functions, in Fig. 1 shown in (a).Transmit letter
The number equation left side is output variable, and right side is the logical expression of input variable and state variable.Such as：output1:=if
(state1==working) then min (input1, capacity) else 0；
9.2 describe with the incidence relation between level function module in system, and (b) is shown in Fig. 1.Connection is found first
The input variable and output variable of two, line both ends function module.Build the logical relation of input variable and output variable.Such as：
Func2.input1:=Func1.output；
9.3 describe the logical relation between function and the input/output variable of subfunction in system, in Fig. 1 shown in (c),
Such as：
function func1
C
E1:subfunc1；E2:subfunc1；E3:subfunc1；
I
input:float:in；
L
E1.input1:=input,
end function
Wherein, E1.input1:=input describes the input of input input1 and father's function func1 of subfunction E1
It is relation of equality；
10. ER represents transmission function in ten tuples, the function box comprising subfunction is needed to describe this, otherwise without
It needs to describe.Transmission function represents the incidence relation of function and the event in subfunction.The incidence relation of event includes three types：
Synchronization association, broadcast association and common cause association；
10.1 structure father's modules with it includes submodule in event synchronization association relationship, in Fig. 2 shown in (a), together
Step association expression event must occur simultaneously, and be had to simultaneously perform by the conversion of event driven state.Representation is ER
{event1&event2……eventn}.Such as：ER { e1＆e2 } represent e1 and e2 occur simultaneously, e1 driving state conversion and
The state conversion of e2 drivings is performed simultaneously；
10.2 structure father's modules with it includes submodule in event broadcast incidence relation, in Fig. 2 shown in (b), extensively
When broadcasting association expression some event generation, notify other that there is the event of broadcast incidence relation therewith, if these event-drivens
The switch condition of state be true, then perform, otherwise do not perform.Representation is ER { event1 | event2 ... eventn }；
10.3 structure father's modules with it includes submodule in event common cause correlating event, (c) institute in Fig. 2
Show, common cause event includes broadcast incidence relation, and in addition to this, multiple events with common incidence relation can be as common
Event equally independently occurs.Common cause event is commonly used in father's mould multiple submodule in the block, they both may be due to itself
Internal error causes to fail, it is also possible to because the mistake of some common cause leads to the situation of effect.Representation for ER event1,
event2……eventn}；
Step 4: the method using discrete event dynamic simulation carries out accident deduction, and record the mistake of failure event generation
Imitate probability；
The failure probability is to sum it up averaging acquisition by Multi simulation running, and by each simulation result, and failure probability is
One approximation, with the increase of simulation times, approximation meeting Step wise approximation actual value, the implementation procedure of single emulation mode is such as
Under：
4.1 initialization
4.1.1 t0 and end time tf at the beginning of setting emulation；
4.1.2, the original state of function representated by each functional block diagram is set, ten yuan of simulation model is deduced by accident dynamic
The value that R is specified in group is initialized；
4.2 put simulation clock TIME=t0
4.3 dynamic update event lists find out accident dynamic this moment and deduce the event for meeting trigger condition in simulation model
E, described to meet the state that trigger condition refers to corresponding to event E to convert the switch condition of T be empty or to be true, then carry out with
Lower two steps operation：
4.3.1 the event for being unsatisfactory for trigger condition this moment is deleted from list of thing；
4.3.2 for the event E for meeting trigger condition this moment found out, for each event, if event
It is not present, then the event is added in list of thing in table, and labeled as newly-increased event；
The list of thing meets the event of trigger condition for storing, and is sky when emulating initial, with the fortune of emulation
Traveling Mobile state updates, the event in list of thing according to time of origin by being early ranked up to evening, when simulation time reaches thing
During the time of origin of first event in part list, simultaneously driving condition changes for event triggering, and then pushes emulation stream
The propulsion of journey；
If 4.4 simulation time TIME >=tf or event table are sky, 4.9 are gone to, otherwise, performs 4.5；
4.5 calculate the time of origin for increasing event in list of thing newly, and the time of origin of event is equal to simulation time and event
The summing value of delay time, specific calculation are：For temporal event, immediately occurred when switch condition meets, event is prolonged
The slow time is 0, and time of origin is exactly simulation time；Event, time of origin are postponed for the set time in time delay event
It is the summation of simulation time and delay time, for the chance event in time delay event, needs to take out probability of happening
Sample assumes that the probability of happening of event obeys being uniformly distributed for [0,1] during sampling, passes through the concrete numerical value that random sampling obtains p, p
∈ [0,1] then acquires the delay time of event, by taking exponential distribution as an example, the distribution letter of exponential distribution for different distributions
Number is p=1-e-λt, after obtaining p, the numerical value of t counter can be released, specific formula is：Similarly, for prestige cloth
You are distributed, and formula is：Other distributions and so on, simulation time is added with t just obtains event
Perform the time；
4.6 are arranged the event in list of thing according to time of origin by early to late, and time of origin is taken out from event table
Earliest event Event, simulation time are advanced to the time of origin of this event, that is, put TIME=tEvent；
4.7 events occur and driving condition changes, and judge whether transformed state is failure state, if losing
Effect state, according to simulation time, calculating failure probability is 1/TIME；
The 4.8 transmission function L in ten tuple of model, the input for the system that the state conversion of calculating causes, output variable
Variation and state transition condition true-false value variation, return perform 4.3；
4.9 emulation terminate.
Step 5: deducing failure state and the failure probability that emulation obtains based on accident dynamic, fault mode and shadow are carried out
Analysis is rung, fills in fault modes and effect analysis table, detailed process is：For the mistake inside the function of being found in simulation process
Effect state carries out fault identification, the fault mode corresponding to failure state is described, if the failure state can result in fault mode
Occur, then increase a record newly in fault modes and effect analysis table, fill in the generation in fault modes and effect analysis table
Code, product or function mark, function, fault mode.And it performs following steps and fills in fault modes and effect analysis table
Other, otherwise handle the Next Failure state, until the processing of all failure states is completed, wherein, fill in single record
When：
5.1 are directed to fault mode, and with reference to the failure state inside mission profile and function, what the system of determining was presently in appoints
Business stage and working method, and be filled up in fault modes and effect analysis table；
5.2 are directed to fault mode, analyzing failure cause, and be filled up in fault modes and effect analysis table；
5.3 are directed to fault mode, and analysis failure influences, including local influence, high one layer are influenced and finally influenced.And
It is filled up in fault modes and effect analysis table；
5.4 according to failure effect analysis (FEA) as a result, determine the severity classification of fault mode, and be filled up to fault mode and shadow
It rings in analysis table；
5.5 analyze according to fault mode reason, failure influence etc. as a result, filling in fault detection method successively.And it is filled up to
In fault modes and effect analysis table；
The interpretations of result such as 5.6 influence according to failure, fault detect improve indemnifying measure, and are filled up to fault mode with influencing
It analyzes in table；
The probability value corresponding with failure state that is calculated in 5.7 acquisition simulation process, and be filled up to fault mode with
In this row of the failure rate of impact analysis table.
Advantages of the present invention is with good effect：
Analysis method mentioned in the present invention had both considered hardware physical failure probability, it is contemplated that the logic inside function
And sequential relationship so that analysis result more true and accurate.
Compared to the method for the static failure probability of traditional calculating, the present invention by model dynamic simulation realize to system into
Row time accessibility emulates, and realizes the rational analysis to time-constrain.
Simulation model is deduced by building the accident of individual feature module in the present invention, and establishes the pass of the association between module
System, a reticular structure is configured to by the state in whole system and event.During model dynamic simulation, the influence of event can
System other modules are traveled to by reticular structure, effectively simulate system dynamically associates behavior, convenient for finding multiple work(
It combination function can fail caused by association.
The simulation model built using this method, the description of model is using accurate certainty numerical value.Convenient for finding
Manually reappear and play back during problem.The accurate positionin of failure event is realized, convenient for finding and Check System failure.
Measured functional block diagram during model construction, meets related national military standard standard, and the model of structure can be used directly
It is calculated in traditional fail-safe analysis and safety indexes, avoids multiplicating modeling.
The verification platform built using this method greatly reduces the threshold that uses of user, easy to spread.In modeling module
The technology of use is the technology that related field practitioner generally grasps, and involved model dynamic simulation method is convenient for computer
Program is realized and simulation, greatly reduces engineering construction difficulty.
Description of the drawings
Fig. 1 is three kinds of logical relations of function transmission function；
Fig. 2 is three kinds of incidence relations of event；
Fig. 3 is elevator system indenture level example；
Fig. 4 is system functional block diagram example；
Fig. 5 is that simulation model describes example；
Fig. 6 is fault modes and effect analysis result example；
Specific embodiment
Make further detailed and deeply description to technical solution of the present invention below with reference to drawings and examples.Below with
For aircraft lifting and landing rudder system, deduced based on accident dynamic the fault modes and effect analysis of emulation, its step are as follows：
Step 1: according to product agreement, product demand and the design documentation of analyzed object, analyzed object is described
Composition and function, according to complete machine, system, subsystem, Field Replaceable Unit, internal field replaceable units level will be analyzed pair
As decomposing layer by layer, need to decompose minimum indenture level according to fault modes and effect analysis；
Description system forms and function is：Elevator system is single-beam box type thin wall structure, and is by beam, girder, rib, illiteracy
Double closed chamber cross-section structures that skin is formed.Operation to ensure elevator system is needed by load, trim performance, is also equipped with counterweight
Trimmer, wing tip counterweight.The results are shown in Figure 3 for description system indenture level；
Step 2：For each function box in functional block diagram, its internal act structure accident dynamic is deduced imitative
True mode.For the trimmer component in Fig. 3, simplified herein and to build system functional block diagram as shown in Figure 4.Work(in Fig. 4
Energy Func includes two subfunctions, is SubFunc1 and SubFunc2 respectively.The incidence relation of three functions is：Function Func's
Input is connected with the input1 of SubFunc1.The output1 of SubFunc1 is connected with the input2 of SubFunc2.SubFunc2
Output2 be connected with the output of Func；
Step 3: for each function box in functional block diagram, its internal act structure accident dynamic is deduced imitative
True mode.Simulation model is a detailed description for including following 10 kinds of elements, including subfunction (C), input variable (I), shape
State variable (S), output variable (O), original state (R), failure state (F), event (E), state conversion (T), transmission function
(L), event correlation relationship (ER), is reduced to<C,I,S,O,R,F,E,T,L,ER>Ten tuples.This sentences the SubFunc1 in Fig. 4
For ten tuples are described, SubFunc1 does not include subfunction, is input1 there are one input variable I, is there are one state variable S
State1, State1 can switch between S1, S2 and S3.SubFunc1 is output1 there are one output variable O.
The original state R of SubFunc1 is S1.The failure state F of SubFunc1 is S3.The four event E difference included in SubFunc1
It is e1, e2, e3, e4.State conversion T in SubFunc1 includes four, is that 1) S1 to S2 is converted to (state1=respectively
S1) [input1==true] |-e1->state1:=S2,2) S2 to S1 is converted to (state1=S2) [input1==
false]|-e2->state1:=S1,3) S3 to S2 is converted to (state1=S3) [input1==false] |-e3->
state1:=S2,4) S1 to S3 is converted to (state1=S1) [input1==false] |-e4->state1:=S3.It is right
The results are shown in Figure 5 after SubFunc2 and Func is also described；
Step 4：Accident deduction is carried out, and record the mistake of failure event generation using the method for discrete event dynamic simulation
Imitate probability.The single simulation algorithm implementation procedure of the corresponding models of Fig. 5 is as follows：
4.1 initialization
4.1.1 the value for putting t0 at the beginning of emulating is 0 and end time tf is 1010Hour；
4.1.2, the original state of function representated by each functional block diagram is set, ten yuan of simulation model is deduced by accident dynamic
The value that R is specified in group is initialized；；The initial value of state1 is S1 in SubFunc1, and state2's is initial in SubFunc2
It is worth for S6.
4.2 put simulation clock TIME=t0, and system just brings into operation at this time, TIME=0.
4.3 dynamic update event lists find out accident dynamic this moment and deduce the event for meeting trigger condition in simulation model
E, described to meet the state that trigger condition refers to corresponding to event E to convert the switch condition of T be empty or to be true.Due to input
Value for true, therefore the value of input1 is also true.And during system initialization, the state of State1 is S1, is not S2, therefore
The value of output1 is false, and the value of input2 is also false.It therefore meets the state conversion of condition includes：S1 to S2's turns
It changes (event e1), S6 to S4's converts (event is e6), and S6 to S5's converts (event is e7).Therefore meet trigger condition
Event E includes e1, e6, e7；
Then following two steps operation is carried out：
4.3.1 the event for being unsatisfactory for trigger condition this moment is deleted from list of thing；List of thing is sky at this time, without deleting
Except event.
4.3.2 for the event E for meeting trigger condition this moment found out, for each event, if event
It is not present, then the event is added in list of thing in table, and labeled as newly-increased event；E1, e6, e7 in list of thing all
It is not present, needs to be added in list of thing, three events are all marked as newly-increased event；
If 4.4 simulation time TIME >=tf or event table are sky, 4.9 are gone to, otherwise, performs 4.5.Simulation time at this time
It is 0, TIME<Tf, list of thing include 3 events e1, e6, e7, are not sky, perform 4.5.Else if TIME >=tf or event
List is empty, performs step 4.9；
4.5 calculate the time of origin for increasing event in list of thing newly.According to the delay function computing relay time of event.e1
Dirac distributions are obeyed, therefore delay time t is equal to parameter 100.E6 obeys exponential distribution, and the calculation formula of delay time isWherein the value of λ is 1E-6.Assuming that it is 0.6 to the value that e6 probability of happening is sampled, according to formulaThe value of delay time t is 3979.40.E7 also obeys exponential distribution, it is assumed that e7 probability of happening p is taken out at random
The value of sample is 0.3.Then the value of delay time t is 154901.96；
4.6 are arranged the event in list of thing according to time of origin by early to late, and time of origin is taken out from event table
Earliest event Event, simulation time are advanced to the time of origin of this event, that is, put TIME=tEvent；It is acquired according in 4.5
E1, e6, e7 delay time, according to being ordered as e1 (100), e6 (3979.40), e7 (154901.96) from small to large.To be
System simulation time TIME is set as 100；
4.7 events occur and driving condition changes, and judge whether transformed state is failure state, if losing
Effect state, according to simulation time, calculating failure probability is 1/TIME；Since the delay time of e1 events is minimum, e1 occurs,
State1 is converted to S2 by state S1.Since S2 is not failure state, without calculating failure probability；
The 4.8 transmission function L in ten tuple of model, the input for the system that the state conversion of calculating causes, output variable
Variation and state transition condition true-false value variation, return perform 4.3；According to output1:=if (state1==S2＆＆
Input1==true) then true else false, when S1 becomes S2, the value of output1 becomes true from false,
Therefore the value of input2 becomes true from false.It returns and performs step 4.2；
4.9 emulation terminate.
In 7 systems of Windows installation Visual Studio, 2012 development platforms, calculating is write using C# programming languages
Machine program realizes above-mentioned simulation algorithm, performs above-mentioned simulation process, may finally obtain in this simulation process SubFunc1 into
Enter the time T of failure state S3S3And SubFunc2 enters the time T of failure state S4S4.Computer program is realized herein
Simulation algorithm perform 1,000,000 times, to TS3And TS4Take average acquisitionWithThe each failure shape of acquisition inverted to the time
State probability of happening isFor It is 7.93 × 10-6；
Step 5: deducing failure state and the failure probability that emulation obtains based on accident dynamic, fault mode and shadow are carried out
Analysis is rung, fills in fault modes and effect analysis table, detailed process is：For the mistake inside the function of being found in simulation process
Effect state carries out fault identification, the fault mode corresponding to failure state is described, if the failure state can result in fault mode
Occur, then increase a record newly in fault modes and effect analysis table, fill in the generation in fault modes and effect analysis table
Code, product or function mark, function, fault mode.And it performs following steps and fills in fault modes and effect analysis table
Other, otherwise handle the Next Failure state, until the processing of all failure states is completed.In this example simplified model, needle
To the failure state S3 and S4 inside the function that is found in simulation process, fault identification is carried out to it, description failure state institute is right
The fault mode answered.It is not that each failure state can result in failure to summarize in real system, therefore is first had to point
Analysis system typical fault pattern, then judges whether the failure state can cause the fault mode.In this example, elevator system
Typical fault pattern includes：1) can not be accurately and timely during control surface deflection deflect into specified position；2) left and right elevator can not be protected
Hold synchronous deflection；3) during aircraft flight steady in a long-term, rudder face cannot keep determining equilbrium position；4) block during control surface deflection
Stagnant phenomenon；5) in-flight rudder face judder；6) trimmer can not be deflected normally as required；7) counterweight loosens；8) rudder face structure
Intensity, rigidity requirement are met, not because fatigue, corrosion etc. lead to the damage of its structure.For failure state S3, Electromagnetic Control
Valve fails, and trimmer non-mandatory movement failure occurs, causes trimmer can not be as required when working under unbalanced mode
Normal deflection influences the control stability of aircraft, has major safety risks.For failure state S4, right lock pressurized strut control
Valve fails, and since the component has backup mechanism, after failure, backup component will have an effect, and will not generate shadow to system
It rings；
5.1 are directed to fault mode, and with reference to the failure state inside mission profile and function, what the system of determining was presently in appoints
Business stage and working method, and be filled up in fault modes and effect analysis table.Non- instruction caused by solenoid valve failure S3
Property failure be happened at taking off when wing flap is put down, between in the recent period；
5.2 are directed to fault mode, analyzing failure cause, and be filled up in fault modes and effect analysis table.Electromagnetism is lived
The failure cause of door failure S3 is that the electromagnet of solenoid valve is stuck motionless；
5.3 are directed to fault mode, and analysis failure influences, including local influence, high one layer are influenced and finally influenced.And
It is filled up in fault modes and effect analysis table.The solenoid valve S3 that fails is to lead to trimmer there are non-instructions to local influence
Property movement failure.Influence to high one layer is that aircraft evelvator right trimmer when unbalanced mode manipulates is in balance mould always
Formula state can not be deflected normally as required.It is final to influence that be aircraft taking off and flight between in the recent period is unstable, have great
Security risk.
5.4 according to failure effect analysis (FEA) as a result, determine fault mode severity classification.And it is filled up to fault mode and shadow
It rings in analysis table.The severity rank of fault mode that solenoid valve failure S3 causes is I；
5.5 analyze according to fault mode reason, failure influence etc. as a result, filling in fault detection method successively.And it is filled up to
In fault modes and effect analysis table.The fault detection method of solenoid valve failure S3 is by system call interception to non-equilibrium mould
Formula, whether detection left and right trimmer is by instruction requirement normal work；
The interpretations of result such as 5.6 influence according to failure, fault detect improve indemnifying measure, and are filled up to fault mode with influencing
It analyzes in table.The improvement indemnifying measure of solenoid valve failure S3 is periodic maintenance and replacement；
The probability value corresponding with failure state that is calculated in 5.7 acquisition simulation process, and be filled up to fault mode with
In this row of the failure rate of impact analysis table.According to system dynamic simulation as a result, the failure probability of solenoid valve failure S3
It is 1.56 × 10-6.Finally formed fault modes and effect analysis table results are as shown in Figure 6.
Claims (1)
- A kind of 1. fault modes and effect analysis method that emulation is deduced based on accident dynamic, it is characterised in that：The step of this method It is rapid as follows：Step 1: according to product agreement, product demand and the design documentation of analyzed object, the composition of analyzed object is described And function, according to complete machine, system, subsystem, Field Replaceable Unit, internal field replaceable units level by analyzed object layer Layer decomposes, and needs to decompose minimum indenture level according to fault modes and effect analysis；Step 2: for each module of minimum indenture level, a function box, constructing function block diagram, in function are established In block diagram, a function box represents a function, and function can include subfunction, according to interdepending between function box And hierarchical relationship, it is connected with the line of no arrow between function box, represents the physical relation between each function box, function box Descriptor format be：" function function titlesFeature content definesend function”；Step 3: for each function box in functional block diagram, emulation mould is deduced to its internal act structure accident dynamic Type, which is a detailed description for including following 10 kinds of elements, including subfunction (C), input variable (I), state Variable (S), output variable (O), original state (R), failure state (F), event (E), state conversion (T), transmission function (L), Event correlation relationship (ER), is reduced to<C,I,S,O,R,F,E,T,L,ER>It is as follows to specifically describe method for ten tuples：C represents subfunction in (1) ten tuple, and the function box comprising subfunction is needed to describe this, otherwise without description. The subfunction that function box includes is write out during description successively.The descriptor format of subfunction is：" function title：Function type ".Such as：function func1CE1:subfunc1；E2:subfunc1；E3:subfunc1；end functionWherein func1 contains function E1, E2 and E3 of three subfunc1 types；I represents input variable in (2) ten tuples, and each function box is required for describing this.Each function module includes one Or multiple input variable.Specific form is " name variable:Types of variables:In ", such as：input1:float:in.If variable With fixed initial value, then types of variables could alternatively be variate-value, such as input2:[0,1]:in；S represents state variable in (3) ten tuples, and the box for not including subfunction is needed to describe this, otherwise without description. State variable refers to some condition or situation in the life cycle of function, and state is a series of knot that function performs activities Fruit, each function module is interior comprising one or more state variables, when meeting corresponding conditions, can carry out state conversion.Shape The form of state variable is " name variable:[variate-value]:S ", such as：state1:[working,failed,repair]:s；R represents original state in (4) ten tuples, and the box for not including subfunction is needed to describe this, otherwise without description. Only there are one original state in each functional block diagram, original state R is one kind in S, and R ∈ S are function institutes when emulation starts The state at place.The form of original state is：" Status Name:=state value ", such as state1:=working；F in (5) ten tuples represents failure state, and the box for not including subfunction is needed to describe this, otherwise need not be retouched It states.In each functional block diagram there are one or multiple failure states, failure state be one kind in S, F ∈ S, in simulation process, If reaching the state, record is needed into the time of the state and calculates failure probability of happening.The form of failure state is： " Status Name:=state value ", such as state1:=failed；O represents output variable in (6) ten tuples, and each function box is required for describing this.Each function module includes one Or multiple output variables.The form of output variable is：" name variable:Types of variables:Out ", such as：output1:float: out；E represents event in (7) ten tuples, and the box for not including subfunction is needed to describe this, otherwise without description.Event It can be converted with triggering state, state conversion must satisfy corresponding switch condition, and when the conditions are met, event triggering can just drive Dynamic state conversion.Functioning as event increases delay time for state conversion.It is big to be broadly divided into two for event in the invention Class：It is time delay event and temporal event respectively, wherein time delay event is chance event (band respectively again comprising two classes Have the probability-distribution function of parameter) and set time delay event.Temporal event is also that thing immediately occurred respectively comprising two classes Part and prerequisite event.The form of chance event in delay event is：" event title (delay:=probability-distribution function (failure probability lambda)) ", such as：failure1(delay:=exponential (1E-10)).Fixation in delay event The form of time delay event is:" event title (delay:=Dirac (delay time)) ", such as：failure2(delay: =Dirac (2)).Temporal event need to only specify event title, and the condition of prerequisite event will retouch in state conversion It states, therefore the form of temporal event is：" event title ".Such as：failure3；T represents state conversion in (8) ten tuples, and the box for not including subfunction is needed to describe this, otherwise without description. State conversion is a kind of relationship between two state variables, represents that object will perform certain action in the state of source, and Enter dbjective state when some particular event occurs or certain specific conditions meet.State conversion form be：" state name Claim：Source state value [switch condition is true] |-event title->Status Name:=dbjective state ", wherein switch condition is optional, when When switch condition is empty, event occurs can driving condition conversion.Such as：(state1=working) |-failure-> state1:When=failed represents that failure events occur, the state of state1 is converted to failed by working；L represents transmission function in (9) ten tuples, and each function box is required for describing this.Transmission function describes three kinds of logics Relationship：It is respectively：(9.1) logical relation between the output inside function and input and state variable, in Fig. 1 shown in (a).Transmission function The equation left side is output variable, and right side is the logical expression of input variable and state variable.Such as：output1:=if (state1==working) then min (input1, capacity) else 0；(9.2) with the incidence relation between level function module in description system, (b) is shown in Fig. 1.Connecting line is found first The input variable and output variable of two, both ends function module.Build the logical relation of input variable and output variable.Such as： Func2.input1:=Func1.output；(9.3) logical relation in description system between function and the input/output variable of subfunction, in Fig. 1 shown in (c), example Such as：function func1CE1:subfunc1；E2:subfunc1；E3:subfunc1；Iinput:float:in；LE1.input1:=input,end functionWherein, E1.input1:=input, the input input1 and the input of father's function func1 for describing subfunction E1 are phases Etc. relationships；ER represents transmission function in (10) ten tuples, and the function box comprising subfunction is needed to describe this, otherwise need not Description.Transmission function represents the incidence relation of function and the event in subfunction.The incidence relation of event includes three types：Together Step association, broadcast association and common cause association；(10.1) structure father's module with it includes submodule in event synchronization association relationship, it is synchronous in Fig. 2 shown in (a) Association expression event must occur simultaneously, and be had to simultaneously perform by the conversion of event driven state.Representation is ER {event1&event2……eventn}.Such as：ER { e1＆e2 } represent e1 and e2 occur simultaneously, e1 driving state conversion and The state conversion of e2 drivings is performed simultaneously；(10.2) structure father's module with it includes submodule in event broadcast incidence relation, in Fig. 2 shown in (b), broadcast When association represents that some event occurs, notify other that there is the event of broadcast incidence relation therewith, if these are event driven The switch condition of state is true, then performs, and is not otherwise performed.Representation is ER { event1 | event2 ... eventn }；(10.3) structure father's module with it includes submodule in event common cause correlating event, in Fig. 2 shown in (c), Common cause event includes broadcast incidence relation, and in addition to this, multiple events with common incidence relation can be as common thing Part equally independently occurs.Common cause event is commonly used in father's mould multiple submodule in the block, they both may be due in itself Portion's mistake causes to fail, it is also possible to because the mistake of some common cause leads to the situation of effect.Representation for ER event1, event2……eventn}；Step 4: the method using discrete event dynamic simulation carries out accident deduction, and the failure for recording failure event generation is general Rate；The failure probability is to sum it up averaging acquisition by Multi simulation running, and by each simulation result, and failure probability is one Approximation, with the increase of simulation times, approximation meeting Step wise approximation actual value, the implementation procedure of single emulation mode is as follows：(4.1) it initializesT0 and end time tf at the beginning of (4.1.1) setting emulation；(4.1.2) sets the original state of function representated by each functional block diagram, and ten tuple of simulation model is deduced by accident dynamic The value that middle R is specified is initialized；(4.2) simulation clock TIME=t0 is put(4.3) dynamic update event list finds out accident dynamic this moment and deduces the event E for meeting trigger condition in simulation model, Described to meet the state that trigger condition refers to corresponding to event E to convert the switch condition of T be empty or to be true, is then carried out following Two steps operate：(4.3.1) deletes the event for being unsatisfactory for trigger condition this moment from list of thing；(4.3.2) for the event E for meeting trigger condition this moment that finds out, for each event, if event table In be not present, then the event is added in list of thing, and labeled as newly-increased event；The list of thing meets the event of trigger condition for storing, and is sky when emulating initial, with emulation operation into Mobile state updates, the event in list of thing according to time of origin by being early ranked up to evening, when simulation time arrival event arranges During the time of origin of first event in table, simultaneously driving condition changes, and then pushes simulation flow for event triggering It promotes；(4.4) if simulation time TIME >=tf or event table are sky, 4.9 are gone to, otherwise, performs 4.5；(4.5) time of origin for increasing event in list of thing newly is calculated, the time of origin of event prolongs equal to simulation time with event The summing value of slow time, specific calculation are：For temporal event, immediately occurred when switch condition meets, event delay Time is 0, and time of origin is exactly simulation time；Event is postponed for the set time in time delay event, time of origin is Simulation time and the summation of delay time for the chance event in time delay event, need to be sampled probability of happening, Assume that the probability of happening of event obeys being uniformly distributed for [0,1] during sampling, the concrete numerical value of p, p ∈ are obtained by random sampling [0,1] then acquires the delay time of event, by taking exponential distribution as an example, the distribution function of exponential distribution for different distributions For p=1- θ-λt, after obtaining p, the numerical value of t counter can be released, specific formula is：Similarly, for Weibull Distribution, formula are：Other distributions and so on, simulation time is added with t just obtains holding for event The row time；(4.6) event in list of thing according to time of origin by early to late is arranged, time of origin is taken out most from event table Early event Event, simulation time are advanced to the time of origin of this event, that is, put TIME=tEvent；(4.7) event occurs and driving condition changes, and judges whether transformed state is failure state, if failure State, according to simulation time, calculating failure probability is 1/TIME；(4.8) the transmission function L in ten tuple of model, the input of system that the state conversion of calculating causes, output variable Variation and the variation of state transition condition true-false value, return and perform 4.3；(4.9) emulation terminates.Step 5: failure state and failure probability based on accident dynamic deduction emulation acquisition, carry out fault mode and influence to divide Analysis, fills in fault modes and effect analysis table, detailed process is：For the failure shape inside the function of being found in simulation process State carries out fault identification, describes the fault mode corresponding to failure state, if the failure state can result in fault mode hair It is raw, then increase a record newly in fault modes and effect analysis table, fill in code in fault modes and effect analysis table, Product or function mark, function, fault mode.And it performs following steps and fills in other in fault modes and effect analysis table , the Next Failure state is otherwise handled, until the processing of all failure states is completed, wherein, when filling in single record：(5.1) for fault mode, with reference to the failure state inside mission profile and function, task that the system of determining is presently in Stage and working method, and be filled up in fault modes and effect analysis table；(5.2) it for fault mode, analyzing failure cause, and is filled up in fault modes and effect analysis table；(5.3) for fault mode, analysis failure influences, including local influence, high one layer are influenced and finally influenced.And it fills out It writes in fault modes and effect analysis table；(5.4) according to failure effect analysis (FEA) as a result, determining the severity classification of fault mode, and be filled up to fault mode with influencing It analyzes in table；(5.5) according to analyses such as fault mode reason, failure influences as a result, filling in fault detection method successively.And it is filled up to event In barrier pattern and impact analysis table；(5.6) it is influenced according to failure, the interpretations of result such as fault detect improve indemnifying measure, and are filled up to fault mode and influence point It analyses in table；(5.7) probability value corresponding with failure state being calculated in simulation process is obtained, and is filled up to fault mode and shadow In this row of the failure rate of sound analysis table.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711248522.3A CN108170892B (en) | 2017-11-30 | 2017-11-30 | Fault mode and influence analysis method based on accident dynamic deduction simulation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711248522.3A CN108170892B (en) | 2017-11-30 | 2017-11-30 | Fault mode and influence analysis method based on accident dynamic deduction simulation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108170892A true CN108170892A (en) | 2018-06-15 |
| CN108170892B CN108170892B (en) | 2021-07-16 |
Family
ID=62525072
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711248522.3A Active CN108170892B (en) | 2017-11-30 | 2017-11-30 | Fault mode and influence analysis method based on accident dynamic deduction simulation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108170892B (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109714201A (en) * | 2018-12-19 | 2019-05-03 | 中国电子产品可靠性与环境试验研究所（（工业和信息化部电子第五研究所）（中国赛宝实验室）） | Network System Reliability appraisal procedure, device, computer equipment and storage medium |
| CN110175359A (en) * | 2019-04-23 | 2019-08-27 | 中国电子产品可靠性与环境试验研究所（（工业和信息化部电子第五研究所）（中国赛宝实验室）） | Complication system Safety Modeling Methods and device based on operation flow |
| CN111046575A (en) * | 2019-12-23 | 2020-04-21 | 中国航空工业集团公司沈阳飞机设计研究所 | Method and system for ensuring simulation consistency |
| CN111459130A (en) * | 2020-03-04 | 2020-07-28 | 辽宁工程技术大学 | Method for determining potential possibility of electric system fault |
| CN112464555A (en) * | 2020-11-10 | 2021-03-09 | 北京航空航天大学 | Multi-agent-based dynamic reliability simulation evaluation method for multi-state system |
| CN112947370A (en) * | 2021-02-03 | 2021-06-11 | 华侨大学 | Safety analysis method and system for steam boiler system |
| CN113051726A (en) * | 2021-03-12 | 2021-06-29 | 哈尔滨工程大学 | Dynamic characteristic analysis method based on coupling of discrete dynamic event tree and RELAP5 |
| CN113067726A (en) * | 2021-03-15 | 2021-07-02 | 中国人民解放军国防科技大学 | Network node failure determination method based on double logic layer agents |
| CN115270425A (en) * | 2022-07-08 | 2022-11-01 | 中国电子产品可靠性与环境试验研究所（（工业和信息化部电子第五研究所）（中国赛宝实验室）） | Method and device for processing fault characteristic mapping relation between logic model and physical model |
| CN116341075A (en) * | 2023-03-29 | 2023-06-27 | 江陵县汇邦建筑工程有限公司 | Intelligent fire control design simulation evaluation system based on intelligent building engineering |
| CN116757058A (en) * | 2023-03-31 | 2023-09-15 | 中国船舶集团有限公司第七一九研究所 | Model module-based fault probability analysis and prediction method and device |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7672756B2 (en) * | 1995-06-07 | 2010-03-02 | Automotive Technologies International, Inc. | Vehicle communications using the internet |
| CN101980478A (en) * | 2010-10-13 | 2011-02-23 | 北京星网锐捷网络技术有限公司 | Method and device for detecting and processing equipment failures and network equipment |
| CN102231080A (en) * | 2011-06-28 | 2011-11-02 | 电子科技大学 | Method for analyzing reliability of harmonic gear for spacecraft based on physics of failure |
| CN105930647A (en) * | 2016-04-18 | 2016-09-07 | 北京航空航天大学 | Beam structure non-probabilistic reliability solving method capable of considering multi-failure modes |
| CN106021001A (en) * | 2016-05-26 | 2016-10-12 | 熊猫电子集团有限公司 | Method for reliability modeling and analysis of electronic products with ring backup structures |
| CN106383967A (en) * | 2016-10-21 | 2017-02-08 | 河南科技大学 | Method for predicting reliability of performance parameter of rolling bearing under updating driving of operation data |
-
2017
- 2017-11-30 CN CN201711248522.3A patent/CN108170892B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7672756B2 (en) * | 1995-06-07 | 2010-03-02 | Automotive Technologies International, Inc. | Vehicle communications using the internet |
| CN101980478A (en) * | 2010-10-13 | 2011-02-23 | 北京星网锐捷网络技术有限公司 | Method and device for detecting and processing equipment failures and network equipment |
| CN102231080A (en) * | 2011-06-28 | 2011-11-02 | 电子科技大学 | Method for analyzing reliability of harmonic gear for spacecraft based on physics of failure |
| CN105930647A (en) * | 2016-04-18 | 2016-09-07 | 北京航空航天大学 | Beam structure non-probabilistic reliability solving method capable of considering multi-failure modes |
| CN106021001A (en) * | 2016-05-26 | 2016-10-12 | 熊猫电子集团有限公司 | Method for reliability modeling and analysis of electronic products with ring backup structures |
| CN106383967A (en) * | 2016-10-21 | 2017-02-08 | 河南科技大学 | Method for predicting reliability of performance parameter of rolling bearing under updating driving of operation data |
Non-Patent Citations (1)
| Title |
|---|
| 张海等: "基于功能角色模型的反馈系统故障模式影响分析", 《国防科技大学学报》 * |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109714201B (en) * | 2018-12-19 | 2021-08-06 | 中国电子产品可靠性与环境试验研究所（（工业和信息化部电子第五研究所）（中国赛宝实验室）） | Network system reliability evaluation method and device, computer equipment and storage medium |
| CN109714201A (en) * | 2018-12-19 | 2019-05-03 | 中国电子产品可靠性与环境试验研究所（（工业和信息化部电子第五研究所）（中国赛宝实验室）） | Network System Reliability appraisal procedure, device, computer equipment and storage medium |
| CN110175359B (en) * | 2019-04-23 | 2023-05-12 | 中国电子产品可靠性与环境试验研究所（（工业和信息化部电子第五研究所）（中国赛宝实验室）） | Method and device for modeling security of complex system based on business process |
| CN110175359A (en) * | 2019-04-23 | 2019-08-27 | 中国电子产品可靠性与环境试验研究所（（工业和信息化部电子第五研究所）（中国赛宝实验室）） | Complication system Safety Modeling Methods and device based on operation flow |
| CN111046575A (en) * | 2019-12-23 | 2020-04-21 | 中国航空工业集团公司沈阳飞机设计研究所 | Method and system for ensuring simulation consistency |
| CN111459130B (en) * | 2020-03-04 | 2022-12-20 | 辽宁工程技术大学 | Method for determining potential possibility of electric system fault |
| CN111459130A (en) * | 2020-03-04 | 2020-07-28 | 辽宁工程技术大学 | Method for determining potential possibility of electric system fault |
| CN112464555A (en) * | 2020-11-10 | 2021-03-09 | 北京航空航天大学 | Multi-agent-based dynamic reliability simulation evaluation method for multi-state system |
| CN112947370A (en) * | 2021-02-03 | 2021-06-11 | 华侨大学 | Safety analysis method and system for steam boiler system |
| CN113051726A (en) * | 2021-03-12 | 2021-06-29 | 哈尔滨工程大学 | Dynamic characteristic analysis method based on coupling of discrete dynamic event tree and RELAP5 |
| CN113067726B (en) * | 2021-03-15 | 2023-04-07 | 中国人民解放军国防科技大学 | Network node failure determination method based on double logic layer agents |
| CN113067726A (en) * | 2021-03-15 | 2021-07-02 | 中国人民解放军国防科技大学 | Network node failure determination method based on double logic layer agents |
| CN115270425A (en) * | 2022-07-08 | 2022-11-01 | 中国电子产品可靠性与环境试验研究所（（工业和信息化部电子第五研究所）（中国赛宝实验室）） | Method and device for processing fault characteristic mapping relation between logic model and physical model |
| CN116341075A (en) * | 2023-03-29 | 2023-06-27 | 江陵县汇邦建筑工程有限公司 | Intelligent fire control design simulation evaluation system based on intelligent building engineering |
| CN116341075B (en) * | 2023-03-29 | 2023-10-24 | 燧人消防科技有限公司 | Intelligent fire control design simulation evaluation system based on intelligent building engineering |
| CN116757058A (en) * | 2023-03-31 | 2023-09-15 | 中国船舶集团有限公司第七一九研究所 | Model module-based fault probability analysis and prediction method and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108170892B (en) | 2021-07-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108170892A (en) | A kind of fault modes and effect analysis method that emulation is deduced based on accident dynamic | |
| JP4875661B2 (en) | Aircraft soundness diagnosis apparatus and method, and program | |
| CN109241583B (en) | Human-computer interaction system reliability solving method based on Markov | |
| CN109492256A (en) | A kind of multidirectional searching method of dynamic system states probability mapping matrix | |
| CN104133941A (en) | FTF-based complex product quality loss network analysis method | |
| US11816024B2 (en) | Method of testing a system model | |
| CN106599352A (en) | Reliability analysis method for aircraft fly-by-wire control system | |
| CN106598766B (en) | A kind of state relation dynamic analysing method for IMA Resources Sharing Mechanism | |
| CN109766568A (en) | A kind of space flight human factors analysis Uncertainty Analysis Method of various factors coupling | |
| Li et al. | Discrete event-driven twin workshop modeling and simulation technology | |
| CN108549734A (en) | TFM three-dimensional information stream modeling methods based on systematic functional structrue | |
| Zibaei et al. | Diagnosis of safety incidents for cyber-physical systems: A uav example | |
| CN104699067A (en) | System fault comprehensive declare processing method | |
| CN111199073A (en) | Function-based task failure analysis method | |
| Gao et al. | Design requirements of PHM system fault diagnosis capability | |
| CN110362939B (en) | Monte carlo-based network target range equipment system reliability modeling system and method | |
| Batson | Critical path acceleration and simulation in aircraft technology planning | |
| El Mir et al. | Certification of machine learning algorithms for safe-life assessment of landing gear | |
| Jiang et al. | Application of prognostic and health management in avionics system | |
| Manikar et al. | A hybrid approach of machine learning and expert knowledge for projection of aircraft operability | |
| CN114118758B (en) | Multi-view-based weapon equipment task section modeling method and system | |
| Li | Research on Software Reliability Evaluation Based on Architecture | |
| Kourdali et al. | A systems engineering method for analysis and simulation of standard operating procedures | |
| Miao et al. | Ground proximity warning mission system dynamic safety modeling based on SoTeRiA | |
| Ji et al. | Software-hardware interdependent reliability assessment technique for software-intensive complex systems |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |