基于ICA方法和改进遗传算法的胎儿心电信号分离

摘要

Because of the independent component analysis (ICA) method is easy to strap into local optimal value, and the genetic algorithm has intrinsic implicit parallelism and better global optimization ability, advantage of adaptively adjust the search direction. This paper proposes an improved genetic algorithm, which is used to extract FECG, and can overcome the defects of the ICA algorithm, improve the separation accuracy. Good separation results are achieved through experiment simulation.%由于独立分量分析(ICA)方法易陷入局部极优值，而遗传算法具有内在的隐并行性和更好的全局寻优能力，自适应地调整搜索方向的优势，本文提出一种改进的遗传算法用于胎儿心电信号的提取，可以克服ICA算法的上述缺陷，改善分离精度。通过实验仿真取得了较好的分离结果。