基于Givens变换和二阶振荡W-C-PSO优化的盲源分离算法

针对智能算法在实现盲源分离时容易陷入局部最优且收敛速度缓慢的问题，提出一种基于Givens变换和二阶振荡粒子群优化的盲源分离算法。该算法首先将惯性权重与学习因子两个参数构造函数关系，使之共同调节算法迭代来提高算法的整体性与全局搜索能力；再引入二阶振荡环节增加种群的多样性，这样算法不易陷入局部最优；此外，采用Givens变换将分离矩阵转换成旋转角度表示形式来降低算法的复杂度。仿真表明，该算法能有效实现机械振动信号和语音信号的盲分离，并且相比其他算法具有更快的收敛速度和更好的分离性能。

Blind separation method based on based on Givens transformation and second-order oscillatory W-C-PSO

For the intelligent algorithm, it is easy to fall into local optimum when implementing blind source separation, and the convergence speed is slow, this paper proposed a blind source separation algorithm based on Givens transform and second-order oscillator particle swarm optimization. This algorithm constructed the functional relationship between the inertia weight and the learning factor, and adjusted the algorithm together to improve the algorithm''s overall and global search ability. The second-order oscillation increased the diversity of the population, so the algorithm was not easy to fall into the local optimum. In addition, the algorithm used Givens transformation to convert the separation matrix into a rotation angle representation to reduce the complexity and accelerated the convergence speed. Simulation results show that the algorithm can effectively achieve the blind separation of mechanical vibration signals and speech signals, and has faster convergence speed and better separation performance than other algorithms.