Team Diamond : Team Diamond Jessica Collins
Christie Dyson
Kandice Mantiply
Patricia Marquis
Joe Villarreal
Brian Wendel
Christa Willis
Overview : Overview Reginald Carpenter: Siemens Orlando, FL
Unidentified product experiencing difference in rated performance vs. measured performance.
Problem: which settings (configuration settings, parameters, and model type) have significant effect on Delta X1 and Delta X2 values.
Rationale : Rationale Team Diamond has determined that performing both a qualitative and quantitative analysis with the given data will be of greatest benefit.
Interactions of variables will be seen.
Impact on the disparity between the rated and measured performance of the product will be seen.
Box Plots and scatterplots will be used to demonstrate the relationship.
Methodology Used : Methodology Used Initial Evaluation of data
Scatterplots
Box Plots
Initial Findings:
We needed to explore the relationship between Delta X1 and Delta X2
Multiple Variables Impact negative delta values
Methodology Used : Methodology Used Multiple Models:
Delta X1 versus Delta X2
Delta X1 and Delta X2 versus Model Type
Delta X1 and Delta X2 versus Parameters
Delta X1 and Delta X2 versus Configurations
Analysis and Findings : Analysis and Findings % Delta X1 and %Delta X2 have strong correlation (-.93) % Delta1 = (Delta 1/Rated X1)*100 % Delta2 = (Delta 2/Rated X2)*100
Delta X1 – Model : Delta X1 – Model Model 104 has the greatest impact on Delta X1 negative values as indicated in the box plot below.
Delta X1 - Parameters : Delta X1 - Parameters Parameter 10 and 11 have the strongest correlation with Delta X1.
Parameters 10 and 11 have a negative linear relationship with Delta X1.
Measured X1 tends to be higher than the rated X1 when Parameter 10 and/or 11 increase.
Configurations : Configurations We found that Configuration 2, on the SP setting, contributes to causing, on average, the greatest negative delta X1 values
Configuration 9, on the SB setting also seems to influence large negative deltaX1 values.
Further evidence is needed to make an accurate conclusion.
ConclusionsModel : ConclusionsModel Model type does affect negative delta X1 values
Model 104 has significant impact on Delta X1.
To achieve more positive Delta X1 values – we recommend model TLRI
Conclusions Parameters : Conclusions Parameters Parameters 10 and 11 more significantly impact DeltaX1 values.
A range between 3040 and 3060 for Parameter 10 and a range between 2680 to 2695 to achieve more positive delta values.
ConclusionsConfigurations : ConclusionsConfigurations We recommend setting Configuration 2 to 2P to achieve positive Delta X1 values.
Summary : Summary Model 104, Parameters 10 and 11 and Configurations 2 have the most significant effect on negative DeltaX1 Value.
We recommend that Siemens engineers further analyze these variables in order to improve the performance of Product P.
An expansion of the sample data should be considered which could increase the level of confidence in the relationships identified and also provide additional conclusions on model specific behavior.