CN106342305B - A kind of testability index requiring towards multitask is determined method - Google Patents

Abstract

The invention discloses a kind of testability index towards multitask requirement and determine method, object is to provide a kind of testability index towards multitask requirement and determines method, makes the error of definite product multitask testability index in allowed band. Technical scheme is: first collect related data information, then determine the requirement of product integrated performance index, then determine that each phased mission properties of product of product require and the incidence relation of testability index, and determine corresponding fault detect rate r_FDiWith Percent Isolated r_FIiDeng testability index, finally determine the product test index that multitask requires. Adopt the present invention to meet under the requirement of product integrated performance index guaranteeing, constantly circulate, revise, obtain the product test index request value that the multitask of minimum testability index error ratio requires, meet the requirement of the integrated performance index of system.

Description

A kind of testability index requiring towards multitask is determined method

Technical field

The present invention relates to the Index of testability engineering field, especially one is wanted towards multitaskThe testability index of asking is determined method.

Background technology

Testability refers to that " product can determine that its state (can work, can not work or performance timely and accuratelyDecline) and effectively isolate a kind of design characteristics of its internal fault ". Testability index is to embody productThe parameter request value of testability demand, the design of guide product testability and measurement product test level, orderBefore can target-seeking testability index have more than 100, and fault detect rate (r wherein_FD) and Percent Isolated (r_FI)It is most important index.

Determine that testability index is the first step of testability engineering. Testability index deterministic process is generally dividedBe three steps:

The first step, collects related data, mainly comprises: 1) Integrated using ensures and analyzes product testRequirement and constraint; 2) available new technology and used the test present situation of product; 3) task of product,Instructions for use and the security requirement to test; 4) maintainability and the maintenance program requirement etc. to test.

Second step, sets up the correlation model of above-mentioned related data and testability index.

The 3rd step, by correlation model is carried out to comprehensive trade-off analysis, requires really according to product performance indexDetermine testability index request.

At present, testability index determines that method mainly contains analogy method, trade-off analysis method and broad sense stochastic Petri(refer to " system testing design analysis and checking " P71-78, BJ University of Aeronautics & Astronautics goes out net methods etc.Version society, 2003; " system engineering and electronic technology " P4-7,2002 (5))

(1) analogy method refers to the testability skill of understanding by inquiry the product of developing and usedArt situation, finds out and requires development and design product at aspects such as instructions for use, component characteristic and technical meritsAkin product, taking its testability index as benchmark, with reference to the General Requirements value of current testability, andThe difference of considering this product and similar products, the testability of determining this product after slightly revising refers toMark.

(2) trade-off analysis method refers to by the computing formula of expansion availability and reliability, and fault is examinedSurvey rate and Percent Isolated are as its variable, according to fixed product availability and reliability, by powerWeighing apparatus analysis is determined fault detect rate and Fault Isolation requirement.

(3) Stochastic Petri Nets method by test activity regard as reliability in the whole life cycle of product,An organic component of maintainability activity, surveys by the product of setting up based on Stochastic Petri NetsTry process model, built the incidence relation of testability index and product performance index requirement, meeting productUnder product performance requirement, determine rational testability index requirement.

Although said method can be determined the testability index of product, these methods are all to appoint for oneThe situation of business. For complicated product, generally need to carry out multiple distinct task, each stageProduct configuration, functional requirement and the performance requirement value of task is also different, under different mission requirementses, and instituteMay there is larger difference in the testability index proposing. Too high testability index will bring design riskAnd the wasting of resources, and too low testability index cannot meet performance requirement value. How to thering is multitaskThe product of testbility demand determines that rational testability index is a key issue, also not open at presentDocument relates to the testability index of multitask requirement and determines method.

The performance requirement of product is the key factor of determining product test index request, has multitask and wantsThe product performance demands of asking comprises two aspects: the one, and the properties of product required value of each task phase, twoIt is the combination property required value of product. The product reliability requiring as multitask requires to comprise each taskThe reliability requirement in stage, comprises again the overall reliability requirement of product, and overall reliability requirement is generallyWeighted sum or the product of each task phase reliability requirement calculate.

Owing to existing, task phase is proposed to the inconsistent problem of properties of product required value, the task rank that haveSection is very high to performance requirement value, and some task phase are lower to performance requirement value, makes the comprehensive of productPerformance requirement value is higher than actual demand. How by revising task phase demand, to meet product comprehensiveUnder energy required value, by the lower task phase performance requirement value of raising time, reduce higher appointing as far as possibleBusiness work-in-process performance requirement value, reaches unanimity the performance requirement value of each task phase of product, isA key issue.

Current testability index determines that method is mainly to require to propose for the single task of simple products,Be not suitable for the product that complicated multitask requires; Proposing testability at the multitask product for complicated refers toWhen mark requires, propose respectively the testability index requirement of product for each mission requirements, finally selectLarge testability index requires the final test index request as product. This often causes testability to refer toMark requires too high, and product test design difficulty is strengthened, and affects the flow of research of product, has wastedThe resources such as human and material resources. Be badly in need of the product test index that a kind of effective ways solve multitask requirementRequire irrational problem.

Summary of the invention

The technical problem to be solved in the present invention is: provide a kind of testability index towards multitask requirement trueDetermine method, mission performance is synchronizeed and revised with testability index, make definite product multitask testabilityThe error of index, in allowed band, finally provides a rational product test index request.

Technical scheme of the present invention is as follows:

The 1st step, collects related data information. Comprise:

1.1 collect product phased mission performance requirement value P from product design specification or contract_i, i representsPhased mission sequence number, i=1,2 ..., N, the phased mission sum that N is product, inputs known product rankSection mission performance required value。

1.2 collect each phased mission letter of product from the analysis of product mission requirements and fail-safe analysis resultBreath, comprises task severity B_i, fault rate λ_i, task environment grade E_iWith duration D_i，I=1,2 ..., N, represents phased mission sequence number.

The 2nd step, determines the requirement of product integrated performance index

2.1 calculation stages task importance degrees. Task severity is larger, fault rate is higher, task environment etc.Level is larger and the duration is longer, and the importance degree of task is just larger, the importance degree of each phased mission α₁(i=1,2 ..., N) and adopt aggregative weighted computing method to calculate, calculating formula is:

${\mathrm{\α}}_i=\frac{1}{4}(\frac{B_i}{\underset{j=1}{\overset{N}{\Σ}}{B}_j}+\frac{{\mathrm{\λ}}_i}{\underset{j=1}{\overset{N}{\Σ}}{\mathrm{\λ}}_j}+\frac{E_i}{\underset{j=1}{\overset{N}{\Σ}}{E}_j}+\frac{D_i}{\underset{j=1}{\overset{N}{\Σ}}{D}_j})---\left(1\right)$

2.2 build product integrated performance index P_M：

$P_M=P_1^{{\mathrm{\α}}_1}\·P_2^{{\mathrm{\α}}_2}.....P_N^{{\mathrm{\α}}_N}---\left(2\right)$

Wherein P_iRepresent the performance requirement value of each phased mission, i=1,2 ..., N, represents phased mission orderNumber.

According to the known phased mission performance requirement value of the 1.1st stepCalculating with the 2.2nd step of taskImportance degree α_i, substitution formula (2) obtains the combination property required value of product

$P_M^{*}={\left(P_1^{*}\right)}^{{\mathrm{\α}}_1}\·{\left(P_2^{*}\right)}^{{\mathrm{\α}}_2}.....{\left(P_N^{*}\right)}^{{\mathrm{\α}}_N}---\left(3\right)$

The 3rd step, determines the incidence relation of each phased mission properties of product required value and testability index,And the testability index of definite each phased mission: fault detect rate r_FDiWith Percent Isolated r_FIi。

3.1 set up the phased mission Stochastic Petri Nets model (bibliography that comprises testability index" system engineering and electronic technology " P4-7,2002 (5));

3.2 pairs of Stochastic Petri Nets models solve, and obtain the performance requirement value with phased mission i P_iFor ordinate, with fault detect rate r_FDiOr Percent Isolated r_FIiFor the relation curve of abscissa, wherein P_iWith r_FDiAnd r_FIiProportional (bibliography " system engineering and electronic technology " P4-7,2002 (5));

The 3.3 fault detect rate F at performance requirement value and each phased mission_FDiOr Percent Isolated r_FIi'sOn relation curve, if known stage performance required value, with phased mission performance requirement value P_iSit for verticalHorizontal line of mark picture, the abscissa corresponding with the intersection point of relation curve is corresponding fault detect rate r_FDiWith Percent Isolated r_FIi; If known fault verification and measurement ratio r_FDiWith Percent Isolated r_FIi, taking this value asAbscissa draws a vertical line, and the ordinate corresponding with the intersection point of relation curve is phased mission performance and wantsEvaluation.

The 4th step, determines the product test index of multitask requirement, and method is:

Adopt mission performance synchronize the method for correction with testability index and determine the product survey that multitask requiresExamination property index, step is as follows:

4.1 by the fault detect rate r of product_FDiWith Percent Isolated r_FIiForm a direct index set to be repaired,Fault detect rate value { the r of each phased mission that input is obtained by the 3rd step_FD1，r_FD2，…，r_FDNAnd faultVerification and measurement ratio value { r_FI1，r_FI2，…，r_FIN}。

4.2 extract arbitrary testability index r from direct index set to be repaired_x, for r_xAt phased missionThe required value of i is expressed as r_xi(i=1,2 ..., N), select r_xMaximum in multistage task: rootAccording to phased mission sequence number, by r_xiForm a sequence S_x, then according to ascending order to S_xHeavyNew arrangement obtains sequence S_x′，S_x' N value be r_xMaximum. Hypothetical sequence S_xI positionPut after index rearrangement, be positioned at sequence S_x' k position, define a function g_xBy S_xAnd S_x′In each index be mapped one by one, i.e. g_x(i)=k represents sequence S_xIn i positioning index again arrangeCorresponding S after order_x' in the index of k position, its inverse functionRepresent the rear sequence S of rearrangement_x′In the corresponding sequence S of index of k position_xIn i positioning index, r_xMaximum beThe product bug verification and measurement ratio required value for example with three mission requirementses is sequence S_x＝{r_x1，r_x2，r_x3, it is largeLittle Guan is r_x2＜r_x3＜r_x1, ascending rearrangement is sequence S afterwards_x′＝{r_x2，r_x3，r_x1}＝?Its inverse function isBe that maximum is

4.3 determine the testability index sequence S of product in multitask requirement_xMiddle maximumWith sequence S_xMean value error ratio beta_x：

${\mathrm{\β}}_x=\frac{r_{{\mathrm{xg}}_x^{-1}\left(N\right)}}{\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{r}_{\mathrm{xi}}}\×100\%---\left(4\right)$

The importance degree difference of 4.4 multitasks is larger, the error ratio beta between the testability index allowing_xAlsoLarger, therefore, the error ratio that adopts the variance of the importance degree of multitask to allow as testability indexThat is:

${\mathrm{\β}}_x^{*}=\frac{1}{N}\underset{i=1}{\overset{N}{\Σ}}{({\mathrm{\α}}_i-\frac{1}{N})}^2---\left(5\right)$

WhenTime, represent testability index r_xError in allowed band, testability index r_xValue is relatively consistent, need not revise, and forwards the 4.8th step to, otherwise, forward the 4.5th step to.

4.5 reduce full test indexValue, establish Δ_xRepresent testability index r_xCorrection step-length,Its span is Δ_x∈ [0.1,0.5], revised indexFor:

$r_{{\mathrm{xg}}_x^{-1}\left(N\right)}^{\′}=r_{{\mathrm{xg}}_x^{-1}\left(N\right)}-{\mathrm{\Δ}}_x---\left(6\right)$

4.6 increase minimum testability indexValue, revised indexFor:

$r_{{\mathrm{xg}}_x^{-1}\left(1\right)}^{\′}=r_{{\mathrm{xg}}_x^{-1}\left(1\right)}+{\mathrm{\Δ}}_x---\left(7\right)$

4.7 determine phased mission testability index required value by step 3.3, again obtain theWithThe performance indications of individual phased missionWithThen willWithSubstitution formula (8)Recalculate product integrated performance index P_M：

$P_M={\left(P_{g_x^{-1}\left(1\right)}\right)}^{{\mathrm{\α}}_{g_x^{-1}\left(1\right)}}\·{\left(P_{g_x^{-1}\left(2\right)}\right)}^{{\mathrm{\α}}_{g_x^{-1}\left(2\right)}}.....{\left(P_{g_x^{-1}\left(N\right)}\right)}^{{\mathrm{\α}}_{g_x^{-1}\left(N\right)}}---\left(8\right)$

If 4.7.1And?Forward 4.8 to.

If 4.7.2And?Forward 4.2 to.

If 4.7.3AndForward step 4.6 to.

If 4.7.4And?Forward toStep 4.2.

If 4.8 direct index set to be repaired exist uncorrected index, forward 4.2 to;

4.9 finish.

Adopt the present invention can reach following technique effect:

(1) the present invention in the 2nd step the task severity by collecting, fault rate, task environment etc.The quantization factors such as level and duration, adopt aggregative weighted computing method to obtain phased mission importance degree, and adoptUse based on phased mission performance requirement value and product integrated performance index computational methods all phased missionsPerformance requirement value is carried out COMPREHENSIVE CALCULATING, makes properties of product overall target can reflect each stage of productPerformance requirement value.

(2) the present invention is in the 4th step, by reducing the highest testability index required value, reduce correspondingThe performance requirement value of phased mission, improve the phased mission performance requirement of minimum testability index required valueValue and then improve minimum testability index requirement, is guaranteeing to meet under the requirement of product integrated performance index,Constantly circulate, revise, can obtain the product that the multitask of minimum testability index error ratio requiresTestability index required value, meets the requirement of the integrated performance index of system.

Brief description of the drawings

Fig. 1 is general flow chart of the present invention.

Fig. 2 is the relation curve of the present invention's the 3rd step performance requirement value and fault detect rate, Percent Isolated

Fig. 3 is the flow chart of the present invention's the 4th step.

Specific embodiments

Fig. 1 is general flow chart of the present invention, and step is as follows:

1 collects related data information. Comprise: phased mission performance requirement value, product mission profile number,Severity, fault rate, task environment grade and duration four factors of phased mission, and these fourFactor is with respect to the weight factor of task significance.

2 utilize formula (1) and the requirement of (2) counting yield integrated performance index.

3 utilize Stochastic Petri Nets method to determine each phased mission properties of product required value and testabilityThe incidence relation of index, and the testability index of definite each phased mission of product.

4 methods of utilizing circulation to revise testability index requirement and phased mission performance requirement value are determined manyThe product test index that business requires, then Output rusults. Concrete steps are in conjunction with Fig. 3 introduction.

Fig. 2 is the relation curve of the present invention's the 3rd step performance requirement value and fault detect rate or Percent Isolated

Abscissa in 1 Fig. 2 (a) is fault detect rate, and ordinate is performance requirement value, and curve is propertyEnergy required value and fault detect rate are at the relation curve of phased mission i. Dot (intersection point) on curveCorresponding ordinate is performance requirement value (P_i=0.99), corresponding abscissa is fault detect rate requirement (r_FDi＝0.925)。

Abscissa in 2 Fig. 2 (b) is Percent Isolated, and ordinate is performance requirement value, and curve is propertyEnergy required value and Percent Isolated are at the relation curve of phased mission i. Dot (intersection point) on curveCorresponding ordinate is performance requirement value (P_i=0.99), corresponding abscissa is Percent Isolated requirement (r_FIi＝0.95)。

Fig. 3 is the flow chart of the 4th step in Fig. 1, and step is as follows:

4.1 input product direct index set to be repaired { r_FD，r_FI, and direct index r to be repaired_FDAnd r_FIAt eachThe testability index value of phased mission is { r_FD1，r_FD2，…，F_FDNAnd { r_FI1，r_FI2，…，r_FIN}；

4.2 choose testability index r to be revised_xCorresponding sequence S_xMaximum

4.3 calculate testability index r by formula (4)_xError ratio beta_x：

The 4.4 error ratios that allowed by formula (5) counting yieldAnd error in judgement ratio beta_xWhether littleInIf so, transfer to 4.8 steps, otherwise, forward the 4.5th step to.

4.5 reduce full test index by formula (6)Value be

4.6 increase minimum testability index by formula (7)Value be

4.7 obtain according to step 3.3 the phased mission performance requirement value of revising after testability index againWithAnd substitution formula (8) recalculates product integrated performance index P_M, after then judgement is revisedFull test index and minimum testability index, revise after integrated performance index and target capabilities indexMagnitude relationship:

● whenAnd Forward 4.8 to;

● whenAnd Forward 4.2 to;

● whenAndForward step 4.6 to;

● whenAnd Forward step 4.2 to;

4.8 have judged whether testability index correction, if so, forward 4.2 to, otherwise, forward to 4.9；

The revised full test index request value of 4.9 output, as final result, finishes.

Claims (1)

1. the testability index requiring towards multitask is determined a method, it is characterized in that comprising the following steps:

The 1st step, collects related data information, comprises the following steps:

1.1 collect product phased mission performance requirement value from product design specification or contract, and input is producedProduct phased mission performance requirement valueThe performance requirement value P of each task_iRepresent, i=1,2 ..., N,Represent phased mission sequence number, the phased mission sum that N is product,For known phased mission performance is wantedEvaluation;

1.2 collect each phased mission letter of product from the analysis of product mission requirements and fail-safe analysis resultBreath, comprises task severity B_i, fault rate λ_i, task environment grade E_iWith duration D_i，I=1,2 ..., N, represents phased mission sequence number;

The 2nd step, determine the requirement of product integrated performance index:

2.1 adopt aggregative weighted computing method to calculate the importance degree α of each phased mission_i, calculating formula is:

${\mathrm{\α}}_i=\frac{1}{4}(\frac{B_i}{\underset{j=1}{\overset{N}{\mathrm{\Σ}}}{B}_j}+\frac{{\mathrm{\λ}}_i}{\underset{j=1}{\overset{N}{\mathrm{\Σ}}}{\mathrm{\λ}}_j}\frac{E_i}{\underset{j=1}{\overset{N}{\mathrm{\Σ}}}{E}_j}+\frac{D_i}{\underset{j=1}{\overset{N}{\mathrm{\Σ}}}{D}_j})---\left(1\right)$

2.2 build product integrated performance index P_M：

$P_M=P_1^{{\mathrm{\α}}_1}\·P_1^{{\mathrm{\α}}_1}\mathrm{.....}{P}_N^{{\mathrm{\α}}_N}---\left(2\right)$

According to known phased mission performance requirement valueWith task importance degree α_i, substitution formula (2) obtains productCombination property required value

$P_M^{*}={\left(P_1^{*}\right)}^{{\mathrm{\α}}_1}\·{\left(P_2^{*}\right)}^{{\mathrm{\α}}_2}\mathrm{.....}{\left(P_N^{*}\right)}^{{\mathrm{\α}}_N}---\left(3\right)$

The 3rd step, determines the incidence relation of each phased mission properties of product required value and testability index,And definite corresponding fault detect rate r_FDiWith Percent Isolated r_FIi, step is as follows:

3.1 set up the phased mission Stochastic Petri Nets model that comprises testability index;

3.2 pairs of Stochastic Petri Nets models solve, and obtain the performance requirement value with phased mission i P_iFor ordinate, with fault detect rate r_FDiOr Percent Isolated r_FIiFor the relation curve of abscissa, wherein P_iWith r_FDiAnd r_FIiProportional;

3.3 at performance requirement value and fault detect rate r_FDiOr Percent Isolated r_FIiRelation curve on, ifKnown stage performance required value, with phased mission performance requirement value P_iFor ordinate is drawn a horizontal line,The abscissa corresponding with the intersection point of relation curve is corresponding fault detect rate r_FDiOr Percent Isolated r_FIi; If known fault verification and measurement ratio r_FDiOr Percent Isolated r_FIi, draw one taking this value as abscissa and hang downStraight line, the ordinate corresponding with the intersection point of relation curve is phased mission performance requirement value;

The 4th step, determines the product test index of multitask requirement

Adopt mission performance synchronize the method for correction with testability index and determine the product survey that multitask requiresExamination property index, step is as follows:

4.1 by the fault detect rate r of product_FDiWith Percent Isolated r_FIiForm a direct index set to be repaired,Fault detect rate value { the r of each phased mission that input is obtained by the 3rd step_FD1，r_FD2，…，r_FDNAnd faultIsolation rate value { r_FI1，r_FI2，…，r_FIN}；

4.2 extract arbitrary testability index r from direct index set to be repaired_x, for r_xAt phased missionThe required value of i is expressed as r_xi, select r_xMaximum in multistage task: according to phased mission sequence number,By r_xiForm a sequence S_x, then according to ascending order to S_xRearrange and obtain sequence S '_x， S′_xN value be r_xMaximum; Hypothetical sequence S_xThe rearrangement of i positioning index after,Be positioned at sequence S '_xK position, define a function g_xBy S_xAnd S '_xIn each index one a pair ofShould get up, i.e. g_x(i)=k represents sequence S_xIn corresponding S ' after the rearrangement of i positioning index_xIn kThe index of individual position, its inverse functionRepresent the rear sequence S ' of rearrangement_xIn k positionThe corresponding sequence S of index_xIn i positioning index, r_xMaximum be

4.3 determine the testability index sequence S of product in multitask requirement_xMiddle maximumWith sequence S_xMean value error ratio beta_x：

${\mathrm{\β}}_x=\frac{r_{{\mathrm{xg}}_x^{-1}\left(N\right)}}{\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{r}_{xi}}\×100\%---\left(4\right)$

The 4.4 error ratios that adopt the variance of the importance degree of multitask to allow as testability index?

${\mathrm{\β}}_x^{*}=\frac{1}{N}\underset{i=1}{\overset{N}{\Σ}}{({\mathrm{\α}}_i-\frac{1}{N})}^2---\left(5\right)$

WhenTime, represent testability index r_xError in allowed band, testability index r_xValue is relatively consistent, need not revise, and forwards the 4.8th step to, otherwise, forward the 4.5th step to;

4.5 reduce full test indexValue, revised indexFor:

$r_{{\mathrm{xg}}_x^{-1}\left(N\right)}^{\′}=r_{{\mathrm{xg}}_x^{-1}\left(N\right)}-{\mathrm{\Δ}}_x---\left(6\right)$

Δ_xFor testability index r_xCorrection step-length, span is Δ_x∈[0.1，0.5]；

4.6 increase minimum testability indexValue, revised indexFor:

$r_{{\mathrm{xg}}_x^{-1}\left(1\right)}^{\′}=r_{{\mathrm{xg}}_x^{-1}\left(1\right)}+{\mathrm{\Δ}}_x---\left(7\right)$

4.7 determine phased mission testability index required value by step 3.3, again obtain theWithThe performance indications of individual phased missionWithThen willWithSubstitution formula (8)Recalculate product integrated performance index P_M：

$P_M={\left(P_{g_x^{-1}\left(1\right)}\right)}^{{\mathrm{\α}}_{g_x^{-1}\left(1\right)}}\·{\left(P_{g_x^{-1}\left(2\right)}\right)}^{{\mathrm{\α}}_{g_x^{-1}\left(2\right)}}\mathrm{.....}{\left(P_{g_x^{-1}\left(N\right)}\right)}^{{\mathrm{\α}}_{g_x^{-1}\left(N\right)}}---\left(8\right)$

If 4.7.1And?Forward 4.8 to;

If 4.7.2And?Forward 4.2 to;

If 4.7.3AndForward step 4.6 to;

If 4.7.4And?Forward toStep 4.2;

If 4.8 direct index set to be repaired exist uncorrected index, forward 4.2 to;

4.9 finish.