Optimal Sensor Placement for Dynamic Testing of Large Structures

초록

Analytical methods for sensor placement in inverse problems are presented through singular value decomposition and Gerschgorin theorem. The proposed selection methods are more efficient than sequential elimination. In physical transformations such as system condensation, the convergence criterion requires maximization of the lowest eigenvalue in the subspace of the condensed degrees of freedom. Mathematical procedures apply the eigenvector expansion theorem and the rank of the transformation matrix plays an important role in assessing the solution accuracy. In the present study, more emphasis is put on the linear independence and condition number of the transformation matrix. Numerical investigations have proved that the methods provide good selections for the system transformation.