Blind source mixed signal accelerated separation method in engineering

Abstract

The invention relates to a blind source mixed signal accelerated separation method in engineering. Different kinds of signals have different inherent kurtosis values, so that when a mixed signal reaches the complete separation, each signal reaches the inherent kurtosis value; the kurtosis accumulation quantity of each separation signal reaches a stable value; by aiming at the defect of a blind source signal separation fixed step length natural gradient algorithm that the convergence speed and the steady state error characteristics cannot be met at the same time, the kurtosis accumulation quantity relevant to the separation signal is used for forming a proportional-differential control self-adaption variable step length algorithm; as long as in an allowed initial step length selection range, the acceleration step length can be regulated by the algorithm in a self-adaption way by freely selecting one initial step length value; and as the result, the separation speed can be accelerated, and the separation precision is ensured. The method provided by the invention only involves with the value of the selected step length initial value eta 0; the separation complete speed by the algorithm is higher than the fixed step length separation speed, and is almost identical; the ideal stable babbling errors are realized; and therefore the engineering use values are realized.

Description

technical field [0001] The invention belongs to the processing technology of blind source signal separation, and relates to an accelerated separation method of blind source mixed signals in engineering under the premise of establishing feasible engineering parameters through experiments, especially a method based on the proportional differential control of the kurtosis cumulative amount of separated signals The digital signal processing method for accelerating the separation process with variable step length, the only engineering parameter involved can be selected within a definite value range, and has engineering use value. Background technique [0002] Independent component analysis is currently the mainstream method for blind source separation. There have been many effective algorithms, and these algorithms have different forms, and they can all be classified as LMS (Least Mean Square) algorithms. These algorithms all have a learning rate parameter optimization problem. H...

AI Technical Summary

Problems solved by technology

The part of the integral item I is actually to accumulate the difference between the preset value and the feedback value in time; when the accumulated value reaches a certain value, it will be processed again to avoid oscillation, but there is an obvious lag in the adjustment

Images