Utilizing Existing Knowledge

Kasai guides the user through all steps of DoE. Preconfigured settings provide the ideal foundation for quickly obtaining the optimum result. The process begins by defining the engine’s input parameters as well as the limits within which they are to vary. These limits can be defined in different ways, e.g. based on maps that are already known. This makes it possible to divide up the test space in any chosen way. These settings then serve as the starting point for generating a test plan that is used for measuring the engine on the test bench.

Configuring Experimental Space

Automatic Model Comparison

Once the measurement data is available, Kasai is used for “training” the models. Here, the software automatically looks for the best model approach. Selecting from either a polynomial model, RBF network or Gaussian process model. The user is then able to assess model quality and select models for optimization. Models can be evaluated in graphic form on the basis of various interactive a diagrams and also be exported through the integrated Matlab interface for further use in optimization programs.

Modeling

Integrated Optimization

Kasai’s “OptiMap” plug-in is an integrated solution for optimizing maps from start to finish. It guides the user through a workflow in which all optimization settings can be selected. The plug-in comes with the option of defining several optimization sequences that can be run and repeated. A separate optimization target can also be defined for each operating point within the map.

Benefits for Your Development Process

• Easily accessible use of DoE
• Less time on the test bench
• Utilization of available knowledge and given specifications
• Wide range of optimization capabilities
• Efficiency in calibrating control units
• Tried and tested over many years and continually improved

Model Evaluation