Related Topics: | ||
This topic provides a brief overview of the major analysis, data management and reporting capabilities provided by DOE++.
You can also review an introduction to the Synthesis Platform and a list of what's new in the Synthesis version.
ReliaSoft's DOE++ software facilitates traditional Design of Experiments (DOE) techniques for studying the factors that may affect a product or process in order to identify significant factors and optimize designs. The software also expands upon standard methods by supporting the reliability DOE methodology, which offers a major breakthrough for reliability-related analyses by providing the proper analysis treatment for life data and any other data where censoring occurs.
DOE++ facilitates the design types employed in traditional DOE and also expands upon the traditional methods to support the analysis of life data. The supported design types are:
One Factor Designs (also known as one-way ANOVAs) for determining whether a particular factor has an effect on a specified output or response.
Factorial Designs for determining which factors have a significant effect on the output of the response and identifying interactions between factors.
Two Level Factorial (full and fractional)
Plackett-Burman
General Full Factorial
Taguchi Orthogonal Array (single level and mixed level)
Response Surface Method Designs for studying the quadratic effects of the factors, which makes these designs well-suited to predictive modeling and optimization.
Central Composite
Box-Behnken
Robust Parameter Designs that aim to minimize the variability of the response in spite of noise factors by combining an inner array of control factors with an outer array of noise factors.
Reliability DOE, which is a special category of DOE where traditional experiment design types are combined with reliability analysis methods to investigate the effects of different factors on the life of a unit. You can use this sort of analysis with any of the design types mentioned above.
DOE++ includes a variety of tools to guide you through the process of creating and modifying your design, as well as analyzing data from an experiment that doesn't fit any of the predefined design types.
When none of the pre-defined design types are feasible, the Optimal Design tool allows you to build a custom design that is optimized for your situation.
The design folios include a design evaluation feature that can help you determine how to best modify an existing design before you implement it.
The Multiple Linear Regression Folio and Free Form Folio allow you to quickly analyze existing data to explore the regression relationships between the responses and factors of interest—without having to build a design at all.
The software provides a flexible array of tools for analyzing the experiment results, including:
Analysis of variance (ANOVA) information, which provides an overview of the effects of factor(s) on response(s).
Fit metrics to help assess the validity of the model.
Level-specific information and comparisons for one factor designs.
Breakdowns of the importance of individual factors, interactions and/or groups of effects to output.
Likelihood ratio test results for reliability DOE.
A wide variety of diagnostic, interpretive and predictive plots are available, including:
Level Plots for comparing the responses at different levels of a factor:
Response vs. Level
Level Mean
Life Characteristic
Box Plot
Mean PDFs
Comparison Chart
Effect Plots for evaluating the effects of factors and factorial interactions on a response:
Effect Probability
Scatter Plot
Pareto Chart
Main Effects
Interactions
Interaction Matrix
Term Effect Plot
Cube Plot
Contour Plot
Residual Plots for determining the validity of the model for a response:
Residual Probability
Residual vs. Fitted
Residual vs. Order
Residual vs. Factor
Residual Histogram
Fitted vs. Actual
Diagnostic Plots to assist you in identifying problematic data points and determining the appropriate response transformation (if any):
Leverage
Cook's Distance
Box-Cox Transformation
The 3-dimensional Surface Plot is also available to help you visually investigate how varying the levels of two factors will affect a response.
Make your initial design as inclusive and far-reaching as you like, then refine it until you have exactly the information you need. DOE++ provides separate analysis results for each response that you select to include in the analysis. Excluding a response is as simple as clearing a checkbox. Reducing the model is just as easy. Simply open the Select Terms window and select to include/exclude factors and interactions individually or in groups, and then re-analyze the data set.
DOE++’s Optimization Folio provides a collection of powerful tools that you can use to explore the factor level combinations that produce response values within the limits you specify. This includes the ability to search for the factor level combination that produces the most desirable output, as well as the ability to display all the combinations that keep the output within the specified limits.
DOE++'s Prediction window allows you to enter your own combinations of factor settings to predict the resulting response values, complete with confidence metrics, based on the fitted model.
For designs with more than one replicate, DOE++ allows you to determine the variability of the response(s) across runs and to analyze that standard deviation information. This offers valuable insight into the sources of variation within the experimental data.
With the Simulation Worksheet, you can link an experiment design to a diagram in BlockSim/RENO to obtain simulated response data, which can then be analyzed in DOE++ in order to investigate the effect of one or more settings on the simulation results.
The quality of your data will always be limited by the quality of the measurement system you use. To help you obtain the best data possible, DOE++ includes three tools for measurement systems analysis, which you can use to evaluate individual measurement systems, as well as compare one to another.
You can attach files to a project to keep related information from external data files together with the rest of the analysis.
DOE++ makes it easy to create a complete array of plots and charts to present your analysis graphically. The Plot Setup allows you to completely customize the "look and feel" of plot graphics, while the RS Draw metafile graphics editor provides the option to insert text, draw objects or mark particular points on plot graphics. You can save your plots in a variety of graphic file formats (*.jpg, *.gif, *.png or *.wmf) for use in other documents.
Overlay Plots (formerly called "MultiPlots") allow you to plot the results from multiple data sets together in the same plot. This can be an effective visual tool for many different applications, such as comparing different data sets or analysis methods (e.g., Design A vs. Design B or MLE vs. Rank Regression) or demonstrating the effects of a design change (e.g., Before vs. After).
Side-by-Side Plots allow you to view (and print) multiple plots for a given data set side-by-side. For example, you may want to show the probability, reliability, pdf and failure rate plots for a given analysis together in the same window. Alternatively, you may wish to compare the probability or pdf plots for a given data set when analyzed with different distributions.
If you want to gather all the analysis results of interest in one place, you can create a spreadsheet report. Use the Report Generator window to automatically add different results from a selected analysis, or copy results from any analysis in the project and paste them in manually. Also, since they function like any other spreadsheet in Synthesis, these reports provide complete in-cell formula support, cell references and over 150 built-in functions. You can type functions directly into cells or use the Function Wizard to build and insert functions with ease; and the integrated Chart Wizard guides you through the steps to create your own graphical charts.
© 1992-2015. ReliaSoft Corporation. ALL RIGHTS RESERVED.