Experimental study of random dynamic loads identification based on weighted regularization method

• We propose a random load identification model considering both measurement noise and FRF error.
• We deduce relative error bound formula of the identified load and analysis of the error source.
• We propose a weighted regularization method based on the proper orthogonal decomposition.

In the identification of random dynamic loads acting on the structure, it is necessary to take into account the effect of measuring error of responses and estimation error of frequency response functions (FRFs). For this purpose, a dynamic load identification model was built in which the relative error bound of the identified loads was estimated. To mitigate error propagation and ill-posed problem during the process of identification, a weighted regularization approach based on the proper orthogonal decomposition (POD) was proposed, where the regularization parameter was selected by the GCV method. To verify the feasibility and effectiveness of this method, experiments of loads identification were performed on a plate structure which was randomly excited by three electromagnetic exciters. The results show a good agreement between the identified random dynamic loads and the actually exerted loads.