Abstract: The inverse problem of impact force identification is typically ill-posed, classical methods for impact force identification, such as the Tikhonov regularization method, have limitations in terms of identification accuracy,hyper-parameter determination and sensor selection. In recent years, the emerging sparse regularization methods have provided new ideas for impact force identification. By means of the joint sparse prior of impact forces in space-frequency domain, a novel model of impact force identification with sparse structure constraints is established.The model is solved by the improved orthogonal matching pursuit algorithm, and the simultaneous localization and reconstruction of the impact forces can be realized in the under-determined circumstances of less measurements. The proposed method is compared with the Tikhonovand L1 regularization methods. Numerical and experimental validations show that the proposed method allows for sparse sensor placement, and can monitor nine potential impact locations using three strain gauges, with high accuracy and good anti-noise performanceand outperforms the Tikhonov regularization and the L1 regularization methods.

Key words: impactforce identification, inverse problem, sparse prior, under-determined system, orthogonal matching pursuit