Hierarchical Bayesian Online Crack Detection Based on Parameter Reduced Order Mode
In order to ensure the stability and accuracy of detection,and effectively deal with the correlation and strong nonlinearity of parameters,a hierarchical Bayesian online crack detection method based on parameter reduced mode was proposed.A hierarchical Bayesian filter for recursive input,state and parameter estimation was proposed,which could only output the measured vibration using spatial incompleteness and noise.The parameters sought were regarded as random variables with finite update states.The dynamics of the parameter states were controlled by an improved estimation strategy,which could explore the parameter space and identify the target val-ue.Furthermore,the parameter reduced order model was used to identify the cracks of the proposed method.Finally,the proposed method was simulated by real components in aerospace applications.The results show that the proposed method has high detection accuracy and stability,which proves the effectiveness of the proposed method.
万方数据
维普
