Research On Structural Dynamic Modeling And Parameters Identifying Of Accelerometers

The comprehensive and accurate description of dynamic characteristics of accelerometers plays a very important role for improved the performance of dynamic test system of acceleration and the dynamic measurement accuracy. The dynamic model of accelerometers as one of the main methods for describing the dynamic characteristics of accelerometers is the basic and key to study the dynamic compensation, dynamic decoupling and dynamic error correction of accelerometers. The existing dynamic modeling research methods of accelerometers are based on the one degree of freedom spring-mass-damper system of second-order model, and obtain the parameters of model by such parameter identification methods as least squares algorithm. Such second-order dynamic model of accelerometers has higher accuracy and can describe the dynamic characteristics of accelerometers in low frequency domain range. But the second-order dynamic model is difficult to characterize the dynamic characteristics of accelerometers in high frequency domain range accurately. The problem of comprehensive and accurate description of dynamic characteristics of accelerometers is urgently needed to solve. The high-order structural dynamic model of accelerometers can better describe the dynamic characteristics of accelerometers. Therefore, it has important theoretical meaning and application value to research on the structural dynamic modeling and parameters identifying of accelerometers.In this paper, the second-order model with time delay of accelerometers is built to eliminate the system time delay based on the one degree of freedom spring-mass-damper system of second model structure. The structural dynamic modeling method of accelerometers based on second-order band-pass filter which can effectively solve the problem of the second dynamic model being hard to describe the dynamic characteristics of accelerometers accurately in high frequency domain range is studied. The method takes the second-order pass filter as basic unit to achieve the dynamic model of accelerometers structural ascending order and builds the high-order structural dynamic model of accelerometers. On the basis of ARMAX of accelerometers, a parameter identification method based on prediction error method for the dynamic model of accelerometers is proposed. And combining the absolute shock dynamic calibration experiment data of accelerometers, the parameters of second-order dynamic model of accelerometers and pure time delay parameter are obtained by the parameter identification method. Then it can align the experiment data by using the pure time delay parameter, and achieve the parameters of second-order band-pass filter of higher-order structural dynamic model of accelerometers combining the experimental data after alignment.The experimental results showed that the dynamic model parameter identification method of accelerometers based on PEM can effectively realize the parameter identification of dynamic model of accelerometers and the second-order dynamic model has higher accuracy which can reflect the dynamic characteristics of accelerometers in low frequency domain range. Compared with second-order dynamic model, the higher-order structural dynamic model which is built by using the structural modeling method of accelerometers can better describe the dynamic characteristics of accelerometers.