The data analysis of GMA multi-physics parameters based on local mean decomposition

An improved local mean decomposition method is presented to compute the data of GMA signals, aiming at the non-stationary and non-linear time-varying characteristics of the GMA signals. The spline interpolation function method is used to compute local mean functions and envelope functions instead of smooth moving average method, which increases the accuracy and convergence rate of the algorithm. The end points of the signal are extended with extreme values in order to decrease the end effect phenomenon in the process of decomposition. Improved algorithm is compared with the EMD decomposition, and the restraining of the end effect phenomenon is analyzed by root mean square values, and decomposition rate is assessed by the decomposition time of the algorithm. The results show that improved LMD algorithm is prior to EMD decomposition either in decomposition rate or in restraining of end effect phenomenon, which is more acceptable for the data computation of GMA signals.