CN104750979A - Comprehensive risk priority number calculating method for architecture - Google Patents
Comprehensive risk priority number calculating method for architecture Download PDFInfo
- Publication number
- CN104750979A CN104750979A CN201510112333.8A CN201510112333A CN104750979A CN 104750979 A CN104750979 A CN 104750979A CN 201510112333 A CN201510112333 A CN 201510112333A CN 104750979 A CN104750979 A CN 104750979A
- Authority
- CN
- China
- Prior art keywords
- fuzzy
- represent
- failure mode
- index
- evaluation index
- 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
- 238000000034 method Methods 0.000 title claims abstract description 71
- 238000011156 evaluation Methods 0.000 claims abstract description 76
- 239000011159 matrix material Substances 0.000 claims description 34
- 238000010606 normalization Methods 0.000 claims description 24
- 238000013459 approach Methods 0.000 claims description 23
- 238000004364 calculation method Methods 0.000 claims description 21
- 238000001514 detection method Methods 0.000 claims description 14
- 230000008569 process Effects 0.000 claims description 10
- 238000004422 calculation algorithm Methods 0.000 claims description 8
- 238000012545 processing Methods 0.000 claims description 5
- 230000006870 function Effects 0.000 claims description 3
- 239000000284 extract Substances 0.000 claims description 2
- 238000011058 failure modes and effects analysis Methods 0.000 abstract description 20
- 230000007547 defect Effects 0.000 abstract description 4
- 238000013461 design Methods 0.000 abstract description 4
- 238000005516 engineering process Methods 0.000 abstract description 4
- 238000004458 analytical method Methods 0.000 description 9
- 230000000661 pacemaking effect Effects 0.000 description 9
- 230000000694 effects Effects 0.000 description 7
- 238000006243 chemical reaction Methods 0.000 description 4
- 230000008447 perception Effects 0.000 description 4
- 230000006872 improvement Effects 0.000 description 3
- SZVJSHCCFOBDDC-UHFFFAOYSA-N iron(II,III) oxide Inorganic materials O=[Fe]O[Fe]O[Fe]=O SZVJSHCCFOBDDC-UHFFFAOYSA-N 0.000 description 3
- 238000012913 prioritisation Methods 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 230000004927 fusion Effects 0.000 description 2
- 210000002837 heart atrium Anatomy 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 239000008280 blood Substances 0.000 description 1
- 210000004369 blood Anatomy 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000012790 confirmation Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000008030 elimination Effects 0.000 description 1
- 238000003379 elimination reaction Methods 0.000 description 1
- 230000009931 harmful effect Effects 0.000 description 1
- 230000007257 malfunction Effects 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000004064 recycling Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000036279 refractory period Effects 0.000 description 1
- 238000012502 risk assessment Methods 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000013519 translation Methods 0.000 description 1
- 230000014616 translation Effects 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Landscapes
- Complex Calculations (AREA)
- Other Investigation Or Analysis Of Materials By Electrical Means (AREA)
Abstract
The invention provides a comprehensive risk priority number calculating method for architecture. Three risk factors (the appearance frequency, the severity degree and the discovery index) in FMEA are evaluated in the mode that fuzzy numbers are utilized in software architecture in a leveled mode; the appearance frequency and the discovery index are determined through component complexity and expertise, and the weights of different risk factors are distinguished through the conception of information entropy; through the conceptions of a positive ideal solution and a negative ideal solution in a TOPSIS method, a more reasonable and accurate risk priority number rank is obtained; in this way, a comprehensive software-failing mode risk factor evaluation technology for the architecture is achieved. By means of the comprehensive risk priority number calculating method, at the initial software design stage, developers are helped to find out potential design defects in a system, and therefore the subsequent software development quality is guaranteed.
Description
Technical field
The invention belongs to reliability engineering technique field, especially the credible analysis field of software architecture, be specifically related to failure model and effect analysis method, Fuzzy Calculation is theoretical.
Background technology
In numerous credible attribute, security is particularly important for the quality of software systems, failure model and effect analysis FMEA method, as a kind of important safety analysis means, is more and more subject to people's attention.The history of FMEA can trace back to the end of the forties in last century.In 1949, US military proposed the concept of FMEA first in the article of a section " Procedure forperforming a failure mode, effects and analysis " by name.To the initial stage sixties, FMEA is applied to aerospace engineering by NASA, and uses FMEA first in Apollo task.From then on, FMEA just develops into one can provide the effective ways of fail message for risk management.But traditional FMEA method mainly contains two problems, one is that it cannot the evaluation and grading factor exactly, and another is that it cannot provide a rational failure mode prioritization.Therefore, a lot of researchist proposes the suggestion of improvement.
The people such as Chang are at Chang C L, Liu P H, Wei C C.Failure mode and effects analysis using greytheory [J] .Integrated Manufacturing Systems, 2001, by adopting a kind of method gray theory based on fuzzy logic to obtain RPN in 12 (3): 211-216.The people such as Chen are at Chen J K, Lee Y C.Risk priority evaluated by ANPin failure mode and effects analysis [J] .Quality tools and techniques, 2007, in 11 (4): 1-6, the method for application network analytic approach (ANP) estimates the weight of three risks and assumptions, and uses new risk assessment data to decide the priority improved.The people such as Yang are at Yang Z, Bonsall S, Wang J.Fuzzy rule-based Bayesian reasoning approachfor prioritization of failures in FMEA [J] .Reliability, IEEE Transactions on, 2008, propose the priority that a kind of Bayesian inference method based on fuzzy rule determines failure mode in 57 (3): 517-528.The people such as Wang are at WangY M, Chin K S, Poon G K K, et al.Risk evaluation in failure mode and effects analysis using fuzzyweighted geometric mean [J] .Expert Systems with Applications, 2009, utilize FUZZY WEIGHTED geometric mean to improve traditional FMEA in 36 (2): 1195-1207, and by last fuzzy RPN value de-fuzzy to obtain the rank of failure mode.
Much make great efforts to improve traditional FMEA although people take, the assessment of three risks and assumptions is still accurate not.This is because appraisal procedure just relies on the subjective judgement of expert, and have ignored the objective attribute of system itself.Mackel is at Mackel O.Software FMEA Opportunities and benefits of FMEA in the developmentprocess of software-intensive technical systems'[C] describe relation between the information of software systems component complexity and risks and assumptions (occurrence frequency and detection index) in //Proceedings of 5th International Symposium onProgrammable Electronic Systems in Safety Related Applications.2002, namely when the component of a system is more complicated, the probability that inefficacy occurs is larger, the probability be detected is less, contrary, when the component of a system is simpler, the probability of appearance of losing efficacy is less, and the probability be detected is larger.But he does not propose a kind of assessment strategy of risk of system.Yacounb etc. are at Yacoub S M, Ammar H H.A methodology for architecture-level reliabilityrisk analysis [J] .Software Engineering, IEEE Transactions on, 2002, use dynamic complexity to measure the complexity of define system framework model in 28 (6): 529-547, and propose a kind of risk evaluation system according to severity and complexity.But this method have ignored the information in occurrence frequency and detection index, and both multiplied result simple are only used to assess the risk of this system.
Through retrieval, find the patent of invention of having authorized: title " a kind of HAZAN method of Automatic manual transmission technique ", the patent No. " ZL201110106935.4 ".The calculating of this patent documentation risk coefficients R PN adopts classic method, RPN (C
j)=S (C
j) × O (C
j) × D (C
j), the value of three risks and assumptions is just simply multiplied by the method, does not consider their respective weights, and O, S, D combination of different failure mode may produce identical RPN, but the meaning that they comprise is diverse.The risks and assumptions value of such as two failure modes is respectively (1,9,2) and (9,1,2), they have identical RPN=18, but these two failure modes are diverse, adopt traditional RPN computing method may cause venture analysis qualified teachers source and waste of time, even likely make more crucial failure mode out in the cold.
And the present invention overcomes and gone up above-mentioned defect, have employed a kind of fuzzy TOPSIS algorithm based on entropy power.By the weight utilizing the concept of information entropy to determine each risks and assumptions, thus distinguish the contribution of different risks and assumptions in ordering strategy.Then, by utilizing " positive ideal solution " in fuzzy TOPSIS and " minus ideal result " concept to calculate approach degree.Compared to the RPN method in traditional F MEA, risks and assumptions is simply multiplied more accurately with rationally.
Summary of the invention
For defect of the prior art, the object of this invention is to provide the synthesization risk priority number calculating method that a kind of Architecture-oriented considers entropy power and fuzzy ideal approach degree.Present invention utilizes the software architecture description language for network environment, have employed the appraisal procedure of a kind of combination member complexity and expertise to determine be in danger frequency and the detection index of failure mode; By the information fusion by subjectivity and objectivity two aspect, assessment result is mutually than ever more accurately with credible; Meanwhile, the result of three risks and assumptions is all utilize the Triangular Fuzzy Number in fuzzy theory to represent, better avoids the inaccurate problem of subjective semantic description boundary.
In the sequence of risk priority number, present invention employs a kind of fuzzy TOPSIS algorithm based on entropy power.First, by the weight utilizing the concept of information entropy to determine each risks and assumptions, thus distinguished the contribution of different risks and assumptions in ordering strategy.Then, by utilizing " positive ideal solution " in fuzzy TOPSIS and " minus ideal result " concept to calculate approach degree.Finally, the priority of failure mode is determined according to the result of approach degree.This sort algorithm is more accurate and reasonable compared to the RPN method be simply multiplied by risks and assumptions in traditional F MEA.
According to the synthesization risk priority number calculating method of Architecture-oriented provided by the invention, comprise the steps:
Step 1, employing software architecture description language, carry out high-rise modeling to system and obtain architectural model;
Step 2, based on architectural model, from architectural model, obtain structure attribute information, extract component failure pattern information, thus component failure pattern in certainty annuity, in conjunction with domain expertise knowledge and malfunction elimination report, the reason of failure mode and impact are analyzed;
Step 3, according to architectural model, the Run-time scenario of combination member, consider component call situation, calculate the average complexity of each component, computing formula is as follows:
Wherein, cpx (Com
i) represent the average complexity of i-th component, | S| represents the scene number existed in architectural model, PS
krepresent the probability that a kth scene occurs, cpx
k(Com
i) representing the complexity of i-th component under a kth scene, i represents element number, and 1≤i≤t, t represents component sum, Com
irepresent i-th component; Because the simulation run scene is for different systems, in order to ensure the consistance of scene moving model result under each system, we are normalized component complexity, by above-mentioned calculated component complexity divided by the maximal value in them, try to achieve the normalization average complexity N_cpx (Com of component
i), its computing formula is as follows:
Wherein, N_cpx (Com
i) representing the normalization average complexity of i-th component, j represents element number, Com
jrepresent a jth component, 1≤j≤t;
Step 4, carry out fuzzy evaluation to risks and assumptions, wherein, described risks and assumptions comprises: occurrence frequency Occurrence, find index D etection, order of severity Severity;
The assessment of occurrence frequency and discovery index need based on the normalization average complexity of the calculated component of previous step.
The determination of occurrence frequency grade adopts the conversion strategy shown in Fig. 3:
Set up the look-up table T1 (look-up table such as shown in Fig. 3) for determining occurrence frequency grade, it records the normalization average complexity N_cpx (Com of many group components
i) with domain expert to the fuzzy evaluation value of the occurrence frequency of the failure mode corresponding to the evaluation of the quality of the fault-avoidance measure of failure mode; The fuzzy evaluation value of the occurrence frequency of failure mode is determined according to look-up table T1.
In this strategy, the normalization average complexity N_cpx (Com of what ordinate represented is component
i), N_cpx (Com from top to bottom
i) complexity that represents constantly increases, 1 to represent its complexity in total system the highest.The quality of fault-avoidance measure that what horizontal ordinate represented is, " high (H) ", " in (M) " and " low (L) " from left to right successively, the every a line of form increases from left to right successively, illustrates that the probability losing efficacy and occur is in continuous increase.
The subjective and objective analysis result of comprehensive above two steps, just can obtain the blur estimation value of occurrence frequency.The present invention adopt be Triangular Fuzzy Number (a, b, c) as fuzzy evaluation result, a, b, c represent the interval limit represented in Triangular Fuzzy Number, and this and exact value in the past estimate it is different.Illustrating, is the component of 0.3 to normalization average complexity, if the fault-avoidance measure quality of expert to wherein certain failure mode gives the evaluation of " low ", so the fuzzy evaluation value of the occurrence frequency of this failure mode is (7,8,9).
In like manner, find that the determination of index ranking adopts the conversion strategy shown in Fig. 4:
Set up the look-up table T2 (look-up table such as shown in Fig. 4) for determining to find index ranking, it records the normalization average complexity N_cpx (Com of many group components
i) with domain expert to the fuzzy evaluation value of the discovery index of the failure mode corresponding to the evaluation of the Detection job of failure mode; The fuzzy evaluation value of the discovery index of failure mode is determined according to look-up table T2.
In strategy, the normalization average complexity N_cpx (Com of what ordinate represented is component
i), N_cpx (Com from top to bottom
i) complexity that represents constantly increases, 1 to represent its complexity in total system the highest.What horizontal ordinate represented is the fault of domain expert to failure mode and the Detection job of inefficacy, " high (H) ", " in (M) " and " low (L) " from left to right successively, the every a line of form increases from left to right successively, illustrates that the possibility be detected that lost efficacy constantly reduces.Illustrating, is the component of 0.3 to complexity, if the Detection job of expert to wherein certain failure mode gives the evaluation of " low ", so the fuzzy evaluation value of the detection index of failure mode is (9,10,10), this is also a Triangular Fuzzy Number.
Severity level obtains according to domain expert's marking, and obtains the fuzzy evaluation value of the order of severity of failure mode through following Fuzzy processing:
Table 1 software architecture failure mode order of severity vague marking table
Grade | The order of severity describes | Vague marking |
A | Have a strong impact on the safety of system, and without warning | (9,10,10) |
B | Have a strong impact on the safety of system, but have warning | (8,9,10) |
C | Thrashing has calamitous harm | (7.8.9) |
D | Thrashing has known harm | (6.7.8) |
E | Thrashing has less harm | (5,6,7) |
F | Thrashing is safe from harm | (4,5,6) |
G | System cloud gray model ability obviously declines | (3,4,5) |
H | System cloud gray model ability has less reduction | (2,3,4) |
I | Affect minimum | (1,2,3) |
J | Not impact | (1,1,2) |
Step 5, to adopt based on entropy power and the algorithm calculation risk priority number of fuzzy TOPSIS.
All based on a decision matrix, the generation of decision matrix must be first provided based on the calculating of entropy power and the calculating of fuzzy TOPSIS.
Suppose there be m failure mode FM
i(i=1,2 ..., m) with n evaluation index C
j(j=1,2 ..., n).A jth evaluation index C of each failure mode
j(j=1,2 ..., n) can be obtained by the assessment of multiple expert.If there be k expert to carry out decision-making, utilize the fuzzy number set operating gained in last point, can obtain the result of calculation of each failure mode to various criterion, computing formula is as follows:
Wherein,
representing the assessment average result of i-th failure mode jth evaluation index, is a Triangular Fuzzy Number,
represent that a kth expert is to the assessment result of i-th failure mode jth evaluation index.So, risk priority number sequencing problem is regarded as a fuzzy decision-making problem of multi-objective, and a fuzzy decision matrix D can be expressed as:
And weight vector W=(w
1, w
2.., w
j... w
n);
Wherein, W represents the relative weighting of n evaluation index, w
jrepresent a jth evaluation index C
jrelative weighting; FM
irepresent i-th failure mode, C
jrepresent a jth desired value,
represent the assessment average result of i-th failure mode jth evaluation index, i=1,2 ..., m, j=1,2 ..., n;
First carry out entropy power after obtaining decision matrix D to calculate.Because entropy can not directly process fuzzy decision matrix, so need the fuzzy number in fuzzy decision matrix D to change into exact value.
In the present embodiment, fuzzy number is all the method for expressing adopting Triangular Fuzzy Number.It is represented by three numbers, namely
a
1, a
2and a
3represent the interval limit in Triangular Fuzzy Number method for expressing.To fuzzy evaluation average result x
ij, the following formula of membership calculates, and formula is as follows:
Wherein,
represent Triangular Fuzzy Number
membership, degree of membership in other words;
Here the average integrated approach of grade conventional in the fuzzy theory adopted, by utilizing following formula by a Triangular Fuzzy Number
transform into perfect number x
ij;
Obtaining perfect number x
ijafter, in order to calculate each perfect number x
ijratio, adopt normalized function process
Wherein, p
ijrepresent i-th failure mode jth evaluation index proportion, x
ijwhat represent is that i-th failure mode is to the exact value of a jth evaluation index.The not commensurate of different index or yardstick are changed into same linear module by this processing procedure, ensure that subsequent calculations compares the rationality of work.Next entropy calculating is carried out, the entropy e of a jth evaluation index
jcan be calculated as:
Wherein, k>0, e
j>=0, ln is natural logarithm.If for a jth evaluation index, all x
ij, i=1,2 ..., m is equal, namely
so e
jget maximal value.
Here k=1/lnm is allowed, so 0≤e
j≤ 1.
When determining entropy, also need to calculate diversified index.For a given jth evaluation index, work as x
ijchange more hour, entropy e
jlarger.As all x
ij, i=1,2 ..., m, when being all equal, e
j=max e
j=1.So, x
ijfor the comparison work of failure mode just without any contribution, because everybody is the same.Work as x
ijwhen changing larger, entropy e
jless, the weight of a jth evaluation index should be larger.So, define diversified index g
j, g
j=1-e
j, namely work as g
jtime larger, the weight of a corresponding jth evaluation index should be larger.When after the diversified index obtaining all evaluation indexes, just a jth evaluation index C can be obtained by following formula
jrelative weighting w
j:
TOPSIS calculating is carried out based on the fuzzy decision matrix of normalization.
In order to by the scope of the Triangular Fuzzy Number in fuzzy decision matrix D between [0,1], the method that we utilize linear-scale to convert becomes a comparable yardstick diversified standard translations, and by calculate needed for the fuzzy decision matrix of normalization
be designated as
Wherein
Be defined as follows:
Wherein,
represent the fuzzy decision matrix of normalization
in fuzzy number element, be a Triangular Fuzzy Number, to Triangular Fuzzy Number
adopt two kinds of processing modes, as Triangular Fuzzy Number result x
ijwhen comparing the poorest solution closer to optimum solution, to Triangular Fuzzy Number
employing formula A calculates; When Triangular Fuzzy Number result compares optimum solution closer to the poorest solution, to Triangular Fuzzy Number
adopt 3.14 formulae discovery.X
ij=(a
ij, b
ij, c
ij), a
ijrepresent fuzzy number x
ijin first value, b
ijrepresent fuzzy number x
ijin second value, c
ijrepresent fuzzy number x
ijin the 3rd value,
represent c in all failure mode fuzzy numbers
ijmaximal value,
represent a in all failure mode fuzzy numbers
ijminimum value; J represents a jth evaluation index, and j ∈ B represents and comments situation of Profit (solve optimum solution situation), and j ∈ C represents that situation (close to the poorest solution situation) is lost in evaluation.
Wherein, B is benefited evaluation index, and namely value is the bigger the better; C is assessment of fees index, is namely worth the smaller the better; It should be noted that in the present embodiment, each risks and assumptions of failure mode is larger, and failure mode is more crucial, and the target of the inventive method finds these critical failure patterns exactly.Therefore, each risks and assumptions operates according to benefited evaluation index.
The weight vector W that information entropy method is calculated and the fuzzy decision matrix of normalization
be multiplied, calculate Weighted Fuzzy decision matrix
In order to convenience of calculation, defining variable
represent Weighted Fuzzy decision matrix
in element; Determine " positive ideal solution " and " minus ideal result ".When the value of each risks and assumptions is larger, then explain the situation poorer, the situation that namely each evaluation index is best is (0,0,0), and the poorest situation is (1,1,1).But the evaluating objects of problem is the failure mode finding most critical, namely the most dangerous situation.So positive ideal solution A
*with minus ideal result A
-be defined as follows:
Wherein,
represent the best situation of a jth evaluation index with
represent a jth situation that evaluation index is the poorest, j=1,2 ..., n.
Calculate ideal solution distance.I-th failure mode is to the distance of positive ideal solution
with the distance of minus ideal result
be defined as follows respectively:
represent the optimal situation of a jth evaluation index,
represent a jth evaluation index least ideal situation;
Wherein,
represent the distance between two fuzzy numbers, be specifically calculated as follows:
Wherein,
with
two Triangular Fuzzy Number,
p
1, p
2, p
3represent Triangular Fuzzy Number respectively
value, q
1, q
2, q
3represent Triangular Fuzzy Number respectively
value;
Calculate approach degree.After obtaining each failure mode to the distance of " positive ideal solution " and " minus ideal result ", approach degree is utilized to sort to all failure modes.The approach degree CC of i-th failure mode
icomputing formula as follows:
Wherein,
represent the distance of i-th failure mode to minus ideal result,
represent the distance of i-th failure mode to positive ideal solution.
Obviously, as i-th failure mode FM
iduring close to positive ideal solution away from minus ideal result, its approach degree CC
imore close to 1.That is, when approach degree more levels off to 1, corresponding failure mode is also more crucial, and its priority is also higher.Thus, the risk priority number that we will obtain by reasonable assessment more.
Compared with prior art, the present invention has following beneficial effect:
1, present invention employs architecture description language and carry out abstract architecture, obtain high level modeling, and on architectural model, extend the FMEA attributes such as failure mode, occurrence frequency, detection index, the order of severity, to support that follow-up FMEA method is to the assessment of software systems.
2, on the risks and assumptions assessment technology of Software failure modes, present invention employs the appraisal procedure of a kind of combination member complexity and expertise to determine be in danger frequency and the detection index of failure mode.By the information fusion by subjectivity and objectivity two aspect, assessment result is mutually than ever more accurately with credible.Meanwhile, the result of three risks and assumptions is all utilize the Triangular Fuzzy Number in fuzzy theory to represent, better avoids the inaccurate problem of subjective semantic description boundary.
3, in failure mode prioritization, present invention employs a kind of fuzzy TOPSIS algorithm based on entropy power.First, by the weight utilizing the concept of information entropy to determine each risks and assumptions, thus distinguished the contribution of different risks and assumptions in ordering strategy.Then, by utilizing " positive ideal solution " in fuzzy ideal approach degree (TOPSIS) method and " minus ideal result " concept to calculate approach degree.Finally, the priority of failure mode is determined according to the result of approach degree.This sort algorithm is more accurate and reasonable compared to the RPN method be simply multiplied by risks and assumptions in traditional F MEA.
Accompanying drawing explanation
By reading the detailed description done non-limiting example with reference to the following drawings, other features, objects and advantages of the present invention will become more obvious:
Fig. 1 is the synthesization risk priority number calculating method process flow diagram that Architecture-oriented considers entropy power and fuzzy ideal approach degree.
Fig. 2 is the architecture modeling view of pacemaker system.
Fig. 3 is the conversion policy map of the determination of occurrence frequency grade.
Fig. 4 is the conversion policy map of the determination finding index ranking.
Table 1 is software architecture failure mode order of severity vague marking table.
Table 2 is pacemaker component failure pattern and failure cause and impact.
Table 3 is the component complicated dynamic behaviour result of pacemaker system.
Table 4 is expert's subjective evaluation value table.
Table 5 is fuzzy decision matrix value table.
Table 6 is entropy power method result of calculation table.
Table 7 is fuzzy TOPSIS method result of calculation.
Table 8 is result of calculation contrast tables of synthesization RPN result of calculation and traditional RPN.
Embodiment
Below in conjunction with specific embodiment, the present invention is described in detail.Following examples will contribute to those skilled in the art and understand the present invention further, but not limit the present invention in any form.It should be pointed out that to those skilled in the art, without departing from the inventive concept of the premise, some distortion and improvement can also be made.These all belong to protection scope of the present invention.
Architecture-oriented considers a synthesization risk priority number calculating method for entropy power and fuzzy ideal approach degree, assesses the risks and assumptions of three in FMEA (occurrence frequency, the order of severity and discovery index) by utilizing fuzzy number in software architecture layers level; Wherein, occurrence frequency and discovery index are determined by combination member complexity and expertise, utilize the concept of information entropy to distinguish the weight of different risks and assumptions simultaneously, and utilize " the positive ideal solution " in fuzzy ideal approach degree (TOPSIS) method and " minus ideal result " concept, obtain more reasonable and risk priority number sequence accurately, thus realize a kind of Software failure modes risks and assumptions assessment technology of synthesization Architecture-oriented.Adopt the method, developer can be helped to find the design defect that system is potential at the Software for Design initial stage, thus ensure the quality of later stage software development.
As shown in Figure 1, flow process of the present invention is divided into several step, will provide concrete embodiment below for each step.Below provide a pacemaker (English is Pacemaker) example so that this method to be described.A pacemaker comprises 5 components usually:
● Reed_Switch (RS) switch: before to device programming, a magnetite switch should be closed.The effect of this switch avoids electronic noise to the harmful effect of system.
● Coil_Driver (CD) drives: accept the pulse from device programming device, or sends pulse to device programming device.These pulses can be counted, and are construed as the binary code of 0 and 1.The code (bit level) of these segments can be packaged into large one section of code (byte level), and is transmitted to Communication_Gnome.Whether whether the notice of forward or negative sense can be sent back to scrambler with confirmation equipment is effective by correct programming or order.
● Communicaiton_Gnome (CG) communication device: accept the byte level code from Coil_Driver, these code identifications are become order, and order is sent to atrium and ventricle module.It can return the notice of forward and negative sense to investigate the process of order to Coil_Driver.
● Ventricular_Model (pace-making model VT) and Atrial_Model (pace-making model AR): these two components are operationally quite similar, can pace-making heart and perception heartbeat.Once after pacemaker is programmed, magnetite will leave Reed_Switch.Atrium and ventricle module accept to perform corresponding function from the order of Communicaiton_Gnome.
Software architecture description language is adopted to carry out architecture modeling as shown in Figure 2 to it:
Pacemaker can according to which position of heart perceived or by pace-making carry out programming definition different operational mode.Based on the programming requirement of pacemaker compiling method (NBG Code) and equipment, in this example, define 6 Run-time scenarios:
● Programming scene: the operational mode of programmable device set device.Programmable device sets up the communication with equipment by controlling magnetite switch, pulse signal is passed to the equipment can explaining these pulses.Equipment again returned data notifies that programming is effective or invalid.
● AVI scene: VT component monitoring heart, beat perception not then when heart, AR component pace-making heart, following refractory period starts.
● VVI scene: when the perception of VT component is less than heartbeat, VT pace-making heart.
● AAI scene: when the perception of AR component is less than heartbeat, AR pace-making heart.
● VVT scene: VT component continues pace-making heart.
● AAT scene: AR component continues pace-making heart.
Based on different scenes, component can play different effects, and between component is also different alternately.This example, by analyzing the behavior of component in different scene, have found 7 failure modes, and their producing cause and the impact caused is summed up in table 2.
Table 2 pacemaker component failure pattern and failure cause and impact
In this example, scene probability is respectively: Programming=0.01, AVI=0.29, AAT=0.15, AAI=0.20, VVT=0.15, VVI=0.2.Utilize these scene probability, the complexity of each component is to the normalization result (table 3 last column) of maximum complexity.
Table 3 pacemaker system component complicated dynamic behaviour result
Next, four experts are selected to carry out subjective evaluation to each failure mode.Scoring content comprises the quality of fault-avoidance process, the quality of fault detect process and failure effect harmfulness three aspects, and appraisal result is as shown in table 4.According to risks and assumptions appraisal procedure, fuzzy occurrence frequency can be obtained by the information of combination member complexity and fault-avoidance quality two dimensions, and fuzzy detection index then can be obtained by the information of combination member complexity and fault detect quality.The order of severity can obtain fuzzy evaluation value according to the code of points of table 1.
Table 4 expert subjective evaluation value table
Previous step obtains the fuzzy evaluation value of each failure mode risks and assumptions, can the fuzzy decision matrix shown in table 5.Each element in this matrix is Triangular Fuzzy Number.
Table 5 fuzzy decision matrix value table
Failure mode | Fuzzy occurrence frequency | Fuzzy detection index | The fuzzy order of severity |
FM 1 | (3.00,4.00,5.00) | (4.50,5.50,6.50) | (2.75,3.75,4.75) |
FM 2 | (2.50,3.50,4.50) | (6.50,7.50,8.50) | (2.50,3.50,4.50) |
FM 3 | (4.50,5.50,6.50) | (7.50,8.50,9.50) | (2.25,3.25,4.25) |
FM 4 | (2.50,3.50,4.50) | (7.00,8.00,9.00) | (4.50,5.50,6.50) |
FM 5 | (7.50,8.50,9.25) | (8.00,9.00,9.50) | (7.50,8.50,9.50) |
FM 6 | (6.00,7.00,8.00) | (8.50,9.50,9.75) | (9.00,10.0,10.0) |
FM 7 | (6.00,7.00,8.00) | (9.00,10.0,10.0) | (8.50,9.50,10.0) |
Based on above fuzzy decision matrix value table, the triangle fuzzy value in fuzzy matrix is converted into perfect number, recycling formula (3) normalized obtains accurate decision matrix D
norm:
After obtaining normalized accurate decision matrix, entropy assessment is used to remove the weights calculating each risks and assumptions.First, use the entropy of formula (4) calculation risk factor occurrence frequency, detection index and the order of severity, be designated as e respectively
o, e
d, e
s.Then, diversified index g is calculated
o, g
d, g
s.Finally, the weights of each risks and assumptions can be obtained.The calculated value of each step of entropy power method is as shown in table 6.
Table 6 entropy power method result of calculation table
Below, the TOPSIS method based on entropy power is utilized to go to sort to risk priority number.The present invention orientates the determination of risk priority number as a decision-making problem of multi-objective, and three risks and assumptions are counted as useful index.Therefore by the fuzzy decision matrix normalization shown in table 5, the entropy weight then previous calculations obtained imparting normalization matrix obtains the fuzzy decision matrix D for carrying out TOPSIS sequence
tOPSIS.
Because three risks and assumptions process as useful index, so " positive ideal solution " and " minus ideal result " is set to A by the present invention respectively
*=[(1,1,1); (1,1,1); (1,1,1)] and A
-=[(0,0,0); (0,0,0); (0,0,0)].Next, the Euclidean distance of utilization definition is calculated the distance between each failure mode and ideal solution.The distance that can obtain between failure mode i and " positive ideal solution " is
the distance that can obtain between failure mode i and " minus ideal result " is
finally, the approach degree CC of failure mode i can just be calculated
i, and result is sorted, determine failure mode priority.Fuzzy TOPSIS method final calculation result is as shown in table 7.
Table 7 fuzzy TOPSIS method result of calculation table
The result of calculation of synthesization risk priority number (RPN) result of calculation and traditional F MEA risk priority number (RPN) in table 7 contrasts by we.
The result of calculation contrast table of table 8 synthesization RPN result of calculation and traditional RPN
The comparing result of two kinds of FMEA sort algorithms of table 4-8 display.Be not difficult to find out, the ranking results of most of failure mode in these two kinds of methods is not identical, the 5th that to have only had FM3 to come identical.In the middle of traditional F MEA, the RPN value of some failure mode is identical, comes same priority position.Such as, FM6 and FM7 has identical RPN value to be 630, but corresponding detection index and the order of severity are all different.It is problematic that different two failure modes are placed on identical priority position.Cause the reason of this problem to be, traditional FMEA is just simply multiplied the value of three risks and assumptions and have ignored the weight of the problem risk factor.Obviously, this computing method are very irrational, because certain risk indicator can be over-evaluated or underestimate.The method proposed unlike, the present invention with it utilizes the concept of entropy to obtain the weighted value of three factors, and namely 0.3328,0.0856,0.5816, and utilize fuzzy TOPSIS method to calculate the approach degree of FM6 higher than FM7.Therefore, the priority of FM6 should higher than FM7.Result illustrates, method of the present invention is more more reasonable and effective than traditional method when the different failure mode of process has the situation of identical RPN.In addition, FM5 (RPN=648) is decided to be the failure mode of system most critical by traditional F MEA, and FM6 is decided to be the failure mode of most critical by method of the present invention.This is because method of the present invention is assigned to the higher weight of the risks and assumptions order of severity.Meanwhile, in failure effect actual analysis, FM6 can cause the serious consequence of patient death, ought to be more crucial than FM5.Therefore, FM6 should give the highest priority, and first designer should consider how to eliminate and improvement failure mode FM6 in the design phase.
Above specific embodiments of the invention are described.It is to be appreciated that the present invention is not limited to above-mentioned particular implementation, those skilled in the art can make various distortion or amendment within the scope of the claims, and this does not affect flesh and blood of the present invention.
Claims (4)
1. a synthesization risk priority number calculating method for Architecture-oriented, is characterized in that, comprise
Step 1: adopt software architecture description language, high-rise modeling is carried out to system and obtains architectural model;
Step 2: based on architectural model, obtains structure attribute information from architectural model, extracts component failure pattern information, thus component failure pattern in certainty annuity;
Step 3: the normalization average complexity calculating component;
Step 4: fuzzy evaluation is carried out to the risks and assumptions of failure mode;
Step 5: the weight determining each risks and assumptions, and calculate approach degree.
2. the synthesization risk priority number calculating method of Architecture-oriented according to claim 1, is characterized in that, in described step 3, calculate the average complexity of each component, computing formula is as follows:
Wherein, cpx (Com
i) represent the average complexity of i-th component, | S| represents the scene number existed in architectural model, PS
krepresent the probability that a kth scene occurs, cpx
k(Com
i) representing the complexity of i-th component under a kth scene, i represents element number, and 1≤i≤t, t represents component sum, Com
irepresent i-th component;
Try to achieve the normalization average complexity N_cpx (Com of component
i), computing formula is as follows:
Wherein, N_cpx (Com
i) representing the normalization average complexity of i-th component, j represents element number, Com
jrepresent a jth component, 1≤j≤t.
3. the synthesization risk priority number calculating method of Architecture-oriented according to claim 2, it is characterized in that, in described step 4, fuzzy evaluation is carried out to the risks and assumptions of failure mode, wherein, described risks and assumptions comprises: occurrence frequency Occurrence, discovery index D etection, order of severity Severity;
The determination of occurrence frequency grade adopts the strategy that converts: set up the look-up table T1 for determining occurrence frequency grade, and it records the normalization average complexity N_cpx (Com of many group components
i) with domain expert to the fuzzy evaluation value of the occurrence frequency of the failure mode corresponding to the evaluation of the quality of the fault-avoidance measure of failure mode; The fuzzy evaluation value of the occurrence frequency of failure mode is determined according to look-up table T1;
Find that the determination of index ranking adopts the strategy that converts: set up the look-up table T2 for determining to find index ranking, it records the normalization average complexity N_cpx (Com of many group components
i) with domain expert to the fuzzy evaluation value of the discovery index of the failure mode corresponding to the evaluation of the Detection job of failure mode; The fuzzy evaluation value of the discovery index of failure mode is determined according to look-up table T2;
Severity level obtains according to domain expert's marking, and obtains the fuzzy evaluation value of the order of severity of failure mode through Fuzzy processing.
4. the synthesization risk priority number calculating method of Architecture-oriented according to claim 2, is characterized in that, in described step 5, adopts the algorithm calculation risk priority number based on entropy power and fuzzy TOPSIS:
Generate fuzzy decision matrix:
Suppose there be m failure mode FM
i, i=1,2 ..., m, and n evaluation index C
j, j=1,2 ..., n; A jth evaluation index C of each failure mode
jobtained by the assessment of k expert; Each failure mode is to the assessment average result of different evaluation index, and computing formula is as follows:
Wherein,
representing the assessment average result of i-th failure mode jth evaluation index, is a Triangular Fuzzy Number,
represent that a kth expert is to the assessment result of i-th failure mode jth evaluation index;
Obtain fuzzy decision matrix D:
And weight vector W=(w
1, w
2.., w
j... w
n);
Wherein, W represents the relative weighting of n evaluation index, w
jrepresent a jth evaluation index C
jrelative weighting; FM
irepresent i-th failure mode, C
jrepresent a jth desired value,
represent the assessment average result of i-th failure mode jth evaluation index, i=1,2 ..., m, j=1,2 ..., n;
First carry out entropy power after obtaining decision matrix D to calculate:
Fuzzy number in fuzzy decision matrix D is changed into exact value:
To fuzzy evaluation average result
the following formula of membership calculates, and formula is as follows:
Wherein,
represent Triangular Fuzzy Number
membership, a
1, a
2and a
3representative
triangular Fuzzy Number method for expressing in interval limit;
Utilize following formula by Triangular Fuzzy Number
transform into perfect number x
ij;
Obtaining perfect number x
ijafter, in order to calculate each perfect number x
ijratio, adopt normalized function process:
Wherein, p
ijrepresent i-th failure mode jth evaluation index proportion, x
ijwhat represent is that i-th failure mode is to the exact value of a jth evaluation index;
The entropy e of a jth evaluation index
jbe calculated as:
Wherein, k > 0, e
j>=0, ln is natural logarithm;
Define diversified index g
j, g
j=1-e
j;
A jth evaluation index C is obtained by following formula
jrelative weighting w
j:
TOPSIS calculating is carried out based on the fuzzy decision matrix of normalization;
By the fuzzy decision matrix of normalization needed for calculating
be designated as
wherein
be defined as follows:
(formula A)
(formula B)
Wherein,
represent the fuzzy decision matrix of normalization
in fuzzy number element, be a Triangular Fuzzy Number;
a
ijrepresent Triangular Fuzzy Number
in first value, b
ijrepresent fuzzy number
in second value, c
ijrepresent fuzzy number
in the 3rd value,
represent the Triangular Fuzzy Number of all failure modes
middle c
ijmaximal value,
represent the Triangular Fuzzy Number of all failure modes
middle a
ijminimum value; J represents a jth evaluation index, and j ∈ B represents assessment situation of Profit, and j ∈ C represents that situation is lost in evaluation;
To Triangular Fuzzy Number
adopt two kinds of processing modes, work as Triangular Fuzzy Number
when comparing the poorest solution closer to optimum solution, to Triangular Fuzzy Number
employing formula A calculates; When Triangular Fuzzy Number result compares optimum solution closer to the poorest solution, to Triangular Fuzzy Number
employing formula B calculates;
Weight vector W and the fuzzy decision matrix of normalization
be multiplied, calculate Weighted Fuzzy decision matrix
Wherein,
represent Weighted Fuzzy decision matrix
in m × n element;
Positive ideal solution A
*with minus ideal result A
-be defined as follows:
Wherein,
represent the optimal situation of a jth evaluation index,
represent a jth evaluation index least ideal situation, definition
represent the situation that a jth evaluation index is best, definition
represent a jth situation that evaluation index is the poorest, j=1,2 ..., n;
Calculate ideal solution distance: i-th failure mode is to the distance of positive ideal solution
to the distance of minus ideal result
be defined as follows respectively:
Wherein,
represent the distance between two fuzzy numbers, be specifically calculated as follows:
Wherein,
with
two Triangular Fuzzy Number,
p
1, p
2, p
3represent Triangular Fuzzy Number respectively
interval limit, q
1, q
2, q
3represent Triangular Fuzzy Number respectively
interval limit;
Approach degree is utilized to sort to all failure modes; The approach degree CC of i-th failure mode
icomputing formula as follows:
Wherein, approach degree is higher, then think that corresponding failure mode is more crucial, the priority of this corresponding failure mode is also higher.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510112333.8A CN104750979B (en) | 2015-03-13 | 2015-03-13 | The synthesization risk priority number calculating method of Architecture-oriented |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510112333.8A CN104750979B (en) | 2015-03-13 | 2015-03-13 | The synthesization risk priority number calculating method of Architecture-oriented |
Publications (2)
Publication Number | Publication Date |
---|---|
CN104750979A true CN104750979A (en) | 2015-07-01 |
CN104750979B CN104750979B (en) | 2017-12-15 |
Family
ID=53590655
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510112333.8A Expired - Fee Related CN104750979B (en) | 2015-03-13 | 2015-03-13 | The synthesization risk priority number calculating method of Architecture-oriented |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN104750979B (en) |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105242155A (en) * | 2015-11-18 | 2016-01-13 | 南京工程学院 | Transformer fault diagnosis method based on entropy weight method and grey correlation analysis |
CN105591875A (en) * | 2016-01-13 | 2016-05-18 | 北京理工大学 | Multitask-oriented self-networking dynamic matching selection method |
CN106650945A (en) * | 2016-11-16 | 2017-05-10 | 北京航空航天大学 | Software architecture security evaluation method based on theory of evidence combination |
CN107368693A (en) * | 2017-08-18 | 2017-11-21 | 武汉理工大学 | A kind of industrial equipment health state evaluation method |
CN108510209A (en) * | 2018-04-16 | 2018-09-07 | 北京理工大学 | A kind of process failure pattern-recognition and evaluation method based on fuzzy theory |
CN111126853A (en) * | 2019-12-25 | 2020-05-08 | 华北水利水电大学 | Fuzzy FMEA-based hydraulic engineering risk early warning analysis method and system |
CN111291452A (en) * | 2020-03-06 | 2020-06-16 | 威海海洋职业学院 | Method and system for determining risk of failure mode of ship electric propulsion system |
CN111641595A (en) * | 2020-05-11 | 2020-09-08 | 湖南大学 | Power network security risk assessment method and system |
CN111859808A (en) * | 2020-07-24 | 2020-10-30 | 华北电力大学 | Gas turbine system fault sequencing method based on fuzzy weighted geometric mean |
CN112884320A (en) * | 2021-02-18 | 2021-06-01 | 中铁建华南建设有限公司 | Foundation pit risk assessment method, device, equipment and medium based on entropy model |
CN113610337A (en) * | 2021-06-22 | 2021-11-05 | 中铁建华南建设有限公司 | Foundation pit engineering risk determination method and device, computer equipment and storage medium |
US20220404778A1 (en) | 2021-06-21 | 2022-12-22 | Shenzhen Fulian Fugui Precision Industry Co., Ltd. | Intellectual quality management method, electronic device and computer readable storage medium |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103279640A (en) * | 2013-04-23 | 2013-09-04 | 中国电子科技集团公司第三十六研究所 | Satellite-borne electronic system FMECA method based on rough set theory and reliability prediction |
US20140081442A1 (en) * | 2012-09-18 | 2014-03-20 | Askey Computer Corp. | Product quality improvement feedback method |
CN104200124A (en) * | 2014-09-24 | 2014-12-10 | 南昌航空大学 | Number set concept based process failure risk assessment method |
-
2015
- 2015-03-13 CN CN201510112333.8A patent/CN104750979B/en not_active Expired - Fee Related
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20140081442A1 (en) * | 2012-09-18 | 2014-03-20 | Askey Computer Corp. | Product quality improvement feedback method |
CN103279640A (en) * | 2013-04-23 | 2013-09-04 | 中国电子科技集团公司第三十六研究所 | Satellite-borne electronic system FMECA method based on rough set theory and reliability prediction |
CN104200124A (en) * | 2014-09-24 | 2014-12-10 | 南昌航空大学 | Number set concept based process failure risk assessment method |
Non-Patent Citations (3)
Title |
---|
XIAOYU FU等: "Using Entropy Weight-Based TOPSIS to Implement Failure Mode and Effects Analysis", 《PROCEEDINGS OF THE 6TH ASIA-PACIFIC SYMPOSIUM ON INTERNETWARE ON INTERNETWARE》 * |
伍晓榕 等: "模糊语境下的复杂系统关联FMEA方法", 《浙江大学学报(工学版)》 * |
马存宝 等: "基于模糊理论的FMEA方法研究", 《测控技术》 * |
Cited By (20)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105242155A (en) * | 2015-11-18 | 2016-01-13 | 南京工程学院 | Transformer fault diagnosis method based on entropy weight method and grey correlation analysis |
CN105591875A (en) * | 2016-01-13 | 2016-05-18 | 北京理工大学 | Multitask-oriented self-networking dynamic matching selection method |
CN105591875B (en) * | 2016-01-13 | 2018-12-14 | 北京理工大学 | A kind of ad hoc network Dynamic Matching selection method towards multitask |
CN106650945A (en) * | 2016-11-16 | 2017-05-10 | 北京航空航天大学 | Software architecture security evaluation method based on theory of evidence combination |
CN106650945B (en) * | 2016-11-16 | 2019-02-15 | 北京航空航天大学 | A kind of software architecture security assessment method based on evidence combining theory |
CN107368693A (en) * | 2017-08-18 | 2017-11-21 | 武汉理工大学 | A kind of industrial equipment health state evaluation method |
CN108510209A (en) * | 2018-04-16 | 2018-09-07 | 北京理工大学 | A kind of process failure pattern-recognition and evaluation method based on fuzzy theory |
CN111126853A (en) * | 2019-12-25 | 2020-05-08 | 华北水利水电大学 | Fuzzy FMEA-based hydraulic engineering risk early warning analysis method and system |
CN111126853B (en) * | 2019-12-25 | 2023-08-29 | 华北水利水电大学 | Hydraulic engineering risk early warning analysis method and system based on fuzzy FMEA |
CN111291452B (en) * | 2020-03-06 | 2023-05-26 | 威海海洋职业学院 | Ship electric propulsion system fault mode risk determination method and system |
CN111291452A (en) * | 2020-03-06 | 2020-06-16 | 威海海洋职业学院 | Method and system for determining risk of failure mode of ship electric propulsion system |
CN111641595A (en) * | 2020-05-11 | 2020-09-08 | 湖南大学 | Power network security risk assessment method and system |
CN111859808A (en) * | 2020-07-24 | 2020-10-30 | 华北电力大学 | Gas turbine system fault sequencing method based on fuzzy weighted geometric mean |
CN111859808B (en) * | 2020-07-24 | 2023-11-17 | 华北电力大学 | Gas turbine system fault sequencing method based on fuzzy weighted geometric mean |
CN112884320A (en) * | 2021-02-18 | 2021-06-01 | 中铁建华南建设有限公司 | Foundation pit risk assessment method, device, equipment and medium based on entropy model |
CN112884320B (en) * | 2021-02-18 | 2022-03-29 | 中铁建华南建设有限公司 | Foundation pit risk assessment method, device, equipment and medium based on entropy model |
US20220404778A1 (en) | 2021-06-21 | 2022-12-22 | Shenzhen Fulian Fugui Precision Industry Co., Ltd. | Intellectual quality management method, electronic device and computer readable storage medium |
TWI833098B (en) * | 2021-06-21 | 2024-02-21 | 新加坡商鴻運科股份有限公司 | Intellectual quality management method, electronic device and readable storage medium |
CN113610337A (en) * | 2021-06-22 | 2021-11-05 | 中铁建华南建设有限公司 | Foundation pit engineering risk determination method and device, computer equipment and storage medium |
CN113610337B (en) * | 2021-06-22 | 2024-04-19 | 中铁建华南建设有限公司 | Foundation pit engineering risk determination method, device, computer equipment and storage medium |
Also Published As
Publication number | Publication date |
---|---|
CN104750979B (en) | 2017-12-15 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104750979A (en) | Comprehensive risk priority number calculating method for architecture | |
Li et al. | An ensemble learning-based prognostic approach with degradation-dependent weights for remaining useful life prediction | |
Knezevic et al. | Reliability modelling of repairable systems using Petri nets and fuzzy Lambda–Tau methodology | |
Kaur et al. | Soft computing approaches for prediction of software maintenance effort | |
Sathasivam et al. | Logic mining in neural network: reverse analysis method | |
CN106203481B (en) | Electronic equipment state prediction method based on mixed kernel RVM | |
Ren et al. | Robustness verification for machine-learning-based power system dynamic security assessment models under adversarial examples | |
CN115423318A (en) | Method and system for evaluating health degree of computer equipment based on combination weight | |
Mohsin | Investigating the use of an adaptive neuro-fuzzy inference system in software development effort estimation | |
Zhu et al. | Res-HSA: Residual hybrid network with self-attention mechanism for RUL prediction of rotating machinery | |
Robati et al. | Inflation rate modeling: Adaptive neuro-fuzzy inference system approach and particle swarm optimization algorithm (ANFIS-PSO) | |
Uluskan et al. | Predictive Six Sigma for Turkish manufacturers: utilization of machine learning tools in DMAIC | |
Wang et al. | The criticality of spare parts evaluating model using artificial neural network approach | |
Ebrat et al. | RISK ASSESSMENT OF CONSTRUCTION PROJECTS USING NETWORK BASED ADAPTIVE FUZZY SYSTEM. | |
Yusof et al. | A concise fuzzy rule base to reason student performance based on rough-fuzzy approach | |
Ting et al. | A fuzzy reasoning design for fault detection and diagnosis of a computer-controlled system | |
Ocampo et al. | A framework for capturing uncertainty of group decision-making in the context of the AHP/ANP | |
Iranmanesh et al. | Risk assessment of software projects using fuzzy inference system | |
Yin et al. | A new health state assessment method based on interpretable belief rule base with bimetric balance | |
Kaplan et al. | Fuzzy Reliability Theory in the Decision-Making Process | |
Takács et al. | The AHP extended fuzzy based risk management | |
Norwawi et al. | Recognition decision-making model using temporal data mining technique | |
Jeet et al. | Bayesian and fuzzy approach to assess and predict the maintainability of software: a comparative study | |
Utama et al. | ANFIS multi criteria decision making for overseas construction projects: a methodology | |
Chen et al. | Real-time equipment condition assessment for a class-imbalanced dataset based on heterogeneous ensemble learning |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20171215 |
|
CF01 | Termination of patent right due to non-payment of annual fee |