Abstract: Aiming at the inverse problem of distributed load recovery from strain measurements, we first compare several existing influence coefficient methods and discuss the key problems affecting the load inversion accuracy. To solve the selection problem of unit load cases, a stepwise method based on Schmidt's orthogonalization of the maximum vertical distance is proposed to sequentially select the basis cases from the design load case library, establishing a complete set of engineering feasible distributed load recovery approach. Meanwhile, the Tikhonov regularization method is introduced to solve the ill-conditioning of the inverse matrix. Finally, left wing load rams and optical fiber sensor data of a certain aircraft full-scale fatigue test are adopted for a case study, with 0, 5% and 10% random strain measurement errors introduced, respectively. It is verified that the inverse approach has high prediction accuracy and robustness and is reliable for the flight distributed load identification of structure health monitoring of the new generation aerospace vehicles.

Key words: distributed load recovery, influence coefficient method, Schmidt orthogonality, basis cases, Tikhonov regularization