基于差分进化算法的重力固体潮信号独立分量分析

摘要

为了提取出重力固体潮信号中的独立谐波成分,基于地球自转与月球、太阳相对于地球轨道变化的正交关系建立一个天体间引潮力的分解模型.为了实现对三者所产生的潮汐谐波的分解,将差分进化算法用于独立分量分析中,用以提高整个算法的效率,改进实验的观测精度,从而更有效的得到与该模型相一致的独立成分.实验中,对昆明地区的实测信号进行了处理并与理论值进行了对比,结果表明,该方法可以将重力固体潮信号中各潮汐谐波分量间关系揭示出来且与分解模型相对应,各分量所含的频谱信息与理论值相一致,说明该方法是一种有效的、对重力固体潮信号进行独立分量分析的新方法.%The purpose of this work is to solve the difficulty of extracting orthogonal component of tidal harmonic from gravity earth tide signals.An orthogonal decomposition model for gravity earth tide signal is proposed in this paper.Orthogonal decomposition is implemented by the above model from the aspect of space on the harmonic component of the gravity earth tide signal in order for the extracted harmonic component frequency to reflect the independent geophysical information on space.Thus,a new feature extraction method of gravity earth tide signal is presented by combining independent component analysis(ICA)and differential evolution (DE)algorithm,which is more effective to generate the independent component which is consistent with the gravity solid tidal signal decomposition model.In the experiment,the measured signal is processed and compared with the theoretical value in Kunming area.Experimental result shows that the method is able to reveal the relationship of tidal harmonic component in gravity earth tide signals,which is corresponding to the gravity earth tide signal orthogonal decomposition model.Each component contains spectral information consistent with with theory value.These results indicate that this method is effective and a novle method of orthogonal analysis for gravity solid signal.