Identification of high frequency loads using statistical energy analysis method

The paper proposes an identification method of high frequency loads based on statistical energy analysis theory. The load identification problems of structures subjected to multi excitations are classified into three cases according to the completeness of loads location information and subsystem response set, and the corresponding identification schemes are presented. The approaches for the derivation of load power spectrum from input power are described as well under the assumptions of constant value and linear variation on load power spectrum within each analytical frequency band. The simulated and experimental studies of load identification with the approach were conducted on a plate–shell assembled structural system. The results show that the input power of loads can be identified accurately in three cases using the presented identification method. Besides that, the computed load power spectrums under the constant value assumption have reasonable agreements with the measured load power spectrums.

► The high frequency loads are identified using statistical energy analysis method.
► The load identifications under multi excitations are studied for three cases.
► The load power spectrums are derived from input power under two different assumptions.
► The identification results of a plate–shell assembled structure verify the present approach.