Large Structural Systems With Vibration Signal Processing And Modal Parameter Identification Research

Systems science is one of the greatest scientific achievements of half after last century, also one of the most active in scientific field of science. Based on systems science point of view, as long as meet S=(T,R)'s investigated object can be regarded as systems. Certainly, a large structural system such as bridges, dams can be seen as a system, it can be used systems theory to analyze and solve. Almost all large structure systems generally can not do without vibration, so signal here named vibration signal. The inherent characteristics of large structural systems, both natural frequency, damping ratio and mode shapes, also known as modal parameters of large structure is analyzed based on the dynamic behaviors. Modal Parameter identification is modern method to study the dynamic behavior of large structure, safety in large structure of the system more and more attention today, modal Parameter identification is particularly important.This dissertation proposes to eliminate vibration signal trend of rapid integration approach. Wireless sensor networks collected vibration signals (acceleration signal) obtained by integrating the vibration signals (velocity and displacement signals), the use of polynomial fitting method to eliminate the trend of sampling points generated items, Using the relationship between the coefficient of each order to improve computational speed and accuracy to meet the needs of its discrete wireless sensor network "online" processing, analysis of data requirements. Informed not only by simulation sampling frequency and data length on the accuracy and obtained by finite element modeling and simulation analysis of acceleration signals through the wireless sensor network based on the real actual acceleration signal processing, to further validate and illustrate the effectiveness of the proposed method.Model parameters for large structure has the characteristics of the slow decay, the study authors proposed structure based on mean-value generation time series forecasting model of grey systems, builds a residual correction grey forecasting model, denoted, and its first application large structure in the model parameter forecasting. Systems model is an essential part in contents of systmes science. In practice, it is often very difficult or even impossible to obtain complete model information from investigated object. In such a case, the object is in a situation with "poor" information, and it is very difficult or impossible to solve satisfactorily the problems of a system utilizing traditional theory, whereas it may be expected to get good results if the grey systems theory is adopted. Obviously, the model parameters present of large structure is different from the future, forecasting and forecasting model is the key here. This dissertation proposed the forecasting model of theoretical proof and example calculations show that the model of computation is less, the fitting, forecast satisfactory.By analyzing the free response characteristic roots real and imaginary part of the collocation, can effectively remove the random decrement technique to be free to respond to the noise, get more freedom to respond to the real signal, and thus large structure through the ARMA method for model identification, Noise reduction for the random denoise-ARMA method. And wireless sensor networks on the actual collected vibration signal processing, can prove that the proposed method. Modal parameter identification of large structure systems is in order to determine the health status of the system. To accurately informed the health of systems, this dissertation proposed large structure systems with the natural decay and the unnatural decay, natural decay modal parameter can be get from the grey forecasting model, compare with the modal parameter identify currently.By a set value to determine the health of large structural systems. Module is able to state space and polynomial matrix language the language that unifies the most natural, the most thorough language. This dissertation describes the use of large structure of the system state variable representation of the behavior, analysis of large structure of the system model for modal parameter identification of large structural systems has been a new direction.In the last section,a summary has been given for all the proposed research contents. The innovative points are summarized. Also, the further work has been discussed.