Parameter identification method for dynamic oscillation signal model

Abstract

The present invention discloses a parameter identification method for a dynamic oscillation signal model. When a model parameter is identified, actual constraints on the parameter is considered. The method comprises the steps of: firstly, giving a state estimation value and an initial value of a state estimation error covariance; secondly, within a maximum iteration moment range, obtaining a state predictive value and a predictive error covariance at a next moment by using a prediction step of extended Kalman filter; thirdly, updating the state predictive value and the predictive error covariance at the moment by using a filtration step of the extended Kalman filter, and obtaining the state estimation value and the estimation error covariance at the moment; fourthly, judging whether the state estimation value at the moment meets a corresponding actual constraint condition, and if so, going to the state estimation at the next moment; and if not, performing optimization on the state estimation value at the moment by using an improved particle swarm algorithm, and obtaining an optimal state estimation value that meets the constraint condition at the moment, and on the basis, performing estimation on the state at the next moment.

Description

technical field [0001] The invention relates to a new dynamic oscillation signal model parameter identification method, which belongs to the technical field of signal analysis and parameter identification. Background technique [0002] Generally speaking, the process of studying the composition and eigenvalues ​​of signals is usually called signal analysis. Useful information can only be obtained through necessary analysis and processing of the signal. In practical applications, some signals generated during system operation can provide important information on system stability or other aspects. By analyzing the composition and characteristics of these signals, we can understand the important information transmitted by the signals, and then analyze the working status of the system to ensure the normal operation of the system. When analyzing the signal, the parameters of the mathematical model of the signal are often unknown, and it is necessary to use the method of paramet...

AI Technical Summary

Problems solved by technology

However, due to the structure of the traditional extended Kalman filter itself, it is difficult to directly constrain the constraints on the state vector, so the actual constraints of the parameters to be estimated are mostly ignored in the application process of parameter identification

Images