|
GA User Guide
|
OVERVIEW
The Genetic Algorithm (GA) for Accelerometer Placement is a Matlab-based program that will optimally select accelerometer locations for a modal vibration test. This iterative process replicates the way a new gene sequence is produced through direct reproduction, crossover, and mutation from an existing population. By this process a set of accelerometer locations can be efficiently selected to maximize the linear independence of test-measured mode shapes. The GA can also accommodate multiple FEM configurations, simultaneously selecting the best accelerometer locations for multi-configuration modal tests to minimize the test setup time.
Several key features of the GA are:
- Direct Nastran compatibility—Nastran model can be generated for Genetic Algorithm import using standard Nastran cards, and the optimal accelerometer set can be written directly to Nastran ASET cards
- Works with multiple FEM configurations—Can simultaneously select accelerometer locations on multiple FEM configurations to minimize test set up time for multi-configuration modal tests
- Manual accelerometer location selection—User can manually select accelerometer locations to improve mode shape visualization of test-display models
- Triaxial accelerometer selection—GA can select triaxial accelerometer locations
- Automated—GA can run overnight or in the background in batch mode to examine effects of different numbers of accelerometers
- GUI driven—GUI allows the user full control over the solution parameters and provides continuous feedback on the error progression
- Cost function formulation—Uses industry-standard Pseudo-Orthogonality or Self-MAC cost functions to rank the effectiveness of a set of accelerometer
locations
- Interactively edit selected accelerometers —Allows user to interactively edit the final GA accelerometer set and instantly evaluates adjusted accelerometer set
- Interface with Microsoft Excel through XML—Accelerometer locations, Orthogonality, MAC, and frequency comparisons can be exported to Excel-compatible XML format to document results