Research On Continuous Dynamic Model Parameter Identification Method Of Accelerometer

Accelerometer is important device which makes up high-end power units and shock and vibration measuring instruments,the dynamic model parameter identification is the key and important content to research the dynamic characteristics of the accelerometer.The existing parameter identification method of accelerometer dynamic model are mainly directly using the measured accelerometer input and output data to identify the parameters of accelerometer discrete dynamic model by using least squares,artificial neural network,support vector machine and genetic algorithm methods and so on.And the identified accelerometer dynamic model parameters with good accuracy.However,due to the influence of ambient vibration and the unknown time-delay of measurement channel,directly using the measured accelerometer input and output signals for accelerometer dynamic model parameter identification will cause error.Besides due to the model discretization process will inevitably introduces truncation error and rounding error,which makes error in parameter identification of accelerometer dynamic model.It is difficult to meet the need for high precision application of accelerometer occasions.Compared with the discrete dynamic model of the accelerometer,the parameters of the continuous time dynamic model with more physical meanings.Therefore,it is of great theoretical and practical value to improve the accuracy of parameter identification by study continuous dynamic model parameter identification method of accelerometer,reduce the influence of ambient vibration and measurement channel delay.Based on the analysis of differential equation model of accelerometer,presents a modeling method of accelerometer continuous dynamic model based on modulating function.For the accelerometer input and output data is not synchronized due to the time delay of measurement channel,a continuous dynamic model of accelerometer with time delay is constructed.Proposed an accelerometer continuous dynamic model parameter identification method,and analyzed the influence of ambient vibration and random noise during the dynamic calibration,using independent component analysis to extraction the accelerometer signal.Constructed a parameter identification algorithm based on recursive instrumental variable,realized the accelerometer continuous dynamic model parameter identification by iterative calculation.Simulation experiments and calibration experiments show that the established accelerometer continuous dynamic model with time delay can characterize the dynamic characteristics of the accelerometer and the unknown time delay of measurement channel.The proposed accelerometer continuous dynamic model identification algorithm based on recursive instrumental variable can effectively estimate the parameters.The accuracy and consistency of the parameters of the continuous dynamic model of the accelerometer are improved by independent component extraction and delay compensation.