CAFE (Computer-Aided Formula Environment) is a software package for experimental design, its user-friendliness and Chinese windows will provide users with a quick and convenient way to
*Model construction: Data sets with experimental factors
(independent variables) and quality factors (dependent
variables).
*Modeling the relationship between experimental factors
(independent variables) and quality factors (dependent
variables) by statistical methods.
*Model analysis: Analyze the relationship between experimental
factors (independent variables) and quality factors (dependent
variables).
*Parameter optimization: Find a model that satisfies specific
cons traints (experimental factor, quality factor composition)
and optimizes
Quality improvement solution (combination of experimental
factors) to meet specific objectives (composition of
experimental factors and quality factors, e.g., cost, yield, etc.).
CAFE has the following advantages over traditional DOE
software or Taguchi method software：
• Reliable model validation: Traditional DOE or
Taguchi methods do not distinguish between
in-sample and out-of-sample experimental data.
CAFE uses cross-validation, which can be performed
without the need for a sample.CAFE adopts
cross-validation method, which can achieve the effect of
distinguishing between in-sample and out-of-sample
experimental data without increasing the number of
experiments.CAFE uses cross-validation to distinguish
between in-sample and out-of-sample experimental
data without increasing the number of experiments,
so that the error can be accurately estimated.
• Accurate model building capabilities: Traditional DOE
uses regression analysis.Traditional Taguchi method uses
effect analysis CAFE uses neural-like networks in artificial
in telligence.With the same cross-validation method,
neural-like networks are far more accurate than
regression analysis and Effectiveness analysis is more
accurate.
• Practical quality improvement model: The traditional
DOE mixes quality constraints and objectives.
The traditional DOE method can only deal with a single
quality objective.
The traditional Taguchi method can only deal with a
single quality objective, but in practice, process
performance and product quality often have many
limitations(such as cost and some quality characteristics
up per limit, defect rate and some quality characteristics
lower limit) and objectives CAFE uses the following
quality improvement model to solve these problems.
CAFE uses the following quality improvement models
to solve these problems.
CAFE has several application examples,
including:
including:
- • Taguchi experiment or factor experiment
- • Wire bonding process for copper
- conductor wafer packaging
- • Optimization of CNC tooling machine
- performance
- • Optimization of the mold adhesion of /C packag
- • Optical sheet molding quality and injection
- molding process
- • Experiment on reaction surface method
- • Enzymatic synthesis of hexyl esters for
- optimal adaptation
- • Optimization of hydrofluoric acid reaction
- kiln operation
- • Mixture Design Experiment
- • High performance concrete
- proportioning design
