For on-line monitoring and tracking of the multi-modal vibration signals, It propose a kind of method about simulation vibration signals on-line monitoring and tracking which based on random subspace (SSI) and particle filter (PF) algorithm. SSI algorithm can extract the modal frequency and damping ratio of the vibration system, by the formula of the modal frequency and damping ratio of the vibration system, it can get the system state matrix and output matrix, The state matrix substituted into the equation of state of the output matrix, using PF algorithm for on-line monitoring and tracking of the signals, achieve the noise reduction processing and forecasting analysis of signal. For large-scale machinery, bridges and other structures, It has far-reaching for social and economic development. Using SSI algorithm to extract the modal parameters of the system, further build the state matrix and output matrix of the vibration system, and use PF algorithm conduct for signal’s filtering and prediction, on this basis, It can implement on-line monitoring and early warning control for the state of the structure, trial of the practice inclined cables’s vibration signal turn out that it can provide reliable and stable signal tracking and predictive method.