软件过程质量度量与控制

[e-mail]ƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒ[e-mail]•Measure[ISO/IEC14598-1]CC++•Measure[ISO/IEC14598-1]•Measurement[ISO/IEC14598-1]•Metric[ISO/IEC14598-1]IndicatorIndicatorMetric•Indicator[ISO/IEC14598-1]•ProcessCapability[CMMVersion1.1]•ProcessMaturity[CMMVersion1.1]•ProcessPerformance[CMMVersion1.1][e-mail]––––––2––0-1–––––1–––[e-mail][e-mail]/[e-mail]/[e-mail][e-mail][e-mail]•––•––•––•––••–[e-mail]™––––™––[e-mail]ƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒ[e-mail]•GQMGoal-Question-Measurement•PSMPracticalSoftwareMeasurement“G”“Q”“M”[e-mail]•SG1–SP1.1–SP1.2–SP1.3–SP1.4•SG2–SP2.1–SP2.2–SP2.3–SP2.4•GG2GP2.1–2.10•GG3GP3.1–[e-mail][e-mail][e-mail][e-mail][e-mail]12n…12…m12…[e-mail][e-mail]••••••[e-mail]ƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒ[e-mail]Ø“”Ø–––[e-mail]•Itisconsideredmeaningfulbythecustomer,manager,andperformer.•Itdemonstratedquantifiablecorrelationbetweenprocessperturbationsandbusinessperformance.•Itisobjectiveandunambiguouslydefined.•Itdisplaystrends.•Itisanaturalby-productoftheprocess.•Itissupportedbyautomation.•Metricsmustbesimple,objective,easytocollect,easytointerpret,andhardtomisinterpret.•Metricscollectionmustbeautomatedandnonintrusive,thatis,notinterferewiththeactivitiesofdevelopers.•Metricsmustprovideconsistentassessmentsthroughoutthelifecycle,especiallyinearlyphases,wheneffortstoimprovequalityhaveahighpayoff.•Metrics,bothvaluesandtrends,mustbeusedactivelybymanagementandengineeringpersonnelforcommunicatingprogressandqualityinaconsistentformat.•Metrics’[e-mail]•Themeasuresshouldrelatecloselytotheissueunderstudy.Theseareusuallyissuesofquality,resourceconsumption,orelapsedtime.•Theyshouldhavehighinformationcontent..•Theyshouldpassarealitytest.•Theyshouldpermiteasyandeconomicalcollectionofdata.•Theyshouldpermitconsistentlycollected,well-defineddata.•Theyshouldshowmeasurablevariation.•Theyshouldhavediagnosticvalueandbeabletohelpyouidentifynotonlythatsomething[e-mail]ØØØØØ[e-mail]12n…12…m12…[e-mail][e-mail]Ù–Workandprogress–Budgetedcostandexpenditures–StaffingandteamdynamicsÙ–Changetrafficandstability,whichisrelationshipbetweenopenedandclosedsoftwarechangeorders.–Breakageandmodularity,whichisaveragebreakagetrendovertime–Reworkandadaptability,whichisreworktrendovertime–MTBFandmaturity,[e-mail][e-mail]/[e-mail][e-mail]σσσσσσσσσσσ∇∇σσ∇∇σσ∇∇σ∇σσσσσσ∇σ[e-mail][e-mail]-CMM6061527162620%[e-mail]¾¾[TSP]50%[TSP]50%[TSP]80/[e-mail]“80/20”1.[e-mail]ƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒ[e-mail]Ì–––––Ì––––[e-mail](a)(b)[e-mail]//[e-mail][e-mail][e-mail]•6Sigma•ScatterDiagrams•RunChart•Histograms•Cause-and-effectdiagrams(Fish-Bone)•BarChart•ParetoCharts•RegressionAnalysis•ControlChart(SPC)[e-mail]-3σCL+3σCL-2σCL-σCL+2σCL+σ[e-mail]-R•–––•–n2–RnRjRn10•–X–X2σ–Xσ–[e-mail]-RkjXnXniij≤≤=∑=111Xjknn∑==kiiXkX11XXkjXXRjjj≤≤−=1minmax∑==kiiRkR11RRAXUCLX2+=RAXLCLX2−=32RA=σRDUCLR4=RDLCLR3=XXσΑ[e-mail]••––•–X–X2σ–Xσ–X–“”[e-mail]X-RX-SXmRMAMRcu[e-mail]ÌÌÌÌ[e-mail]123ŠŠŠŠ[e-mail]02468101214161820030405060708090100110120130140150CLUCLLCLLSLUSLτ6σs6σ1σsσ2σs=σzτ0zτ=0zτ03σsσzτ3σs-σz-3σs-σ=τ=3σs-σzτ-3σs-σ[e-mail]ƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒ[e-mail]ØØØCMM/CMMIØØ[e-mail]//////[e-mail]–1021117.3168.75746Project374.512751311Project29.48300411Project102004006008001000120[e-mail]–050100150200250300350()[e-mail]–050100150200250300350400450