Signal extraction method in electromagnetic coupling system based on LM-ABC algorithm

In the electromagnetic coupling system, the sinusoidal coupling signal extraction is crucial. It will be difficult to extract the signal parameters because of the possibility of signal frequency offset and saturation. Hence, this paper proposes an algorithm combined Artificial Bee Colony algorithm (ABC) and the Levenberg-Marquardt(LM) algorithm to extract sinusoidal parameters from saturation sinusoidal signal with frequency deviation. The LM-ABC algorithm take full advantage of the global search ability of ABC Algorithm and the high-precision local optimization capability of LM Algorithm to recover the signal under certain noise conditions and achieve high signal accuracy. Firstly, the unsaturated sinusoidal signal sample is extracted from the sample of the saturated sinusoidal signal. Then, we use the ABC algorithm to calculate an initial value. LM iteration based on the initial value optimizes the signal parameters. The simulation results show that if there are enough sampling points, the method can obtain he accurate parameters at a certain noise level from saturated sinusoidal signal under frequency offset.