【摘要】 煤与瓦斯突出会产生声发射(acoustic emission,AE)信号。针对提取较纯净有效的AE信号问题,提出一种邻域动态调整(D)果蝇算法(fruit fly algorithm,FOA)智能优化粒子滤波(particle filter,PF)的去噪方法。利用果蝇个体表征PF中的每个信号点粒子,优化粒子滤波的重采样过程,并通过动态调整邻域粒子数量来改善果蝇算法的寻优能力和收敛速度。以均方根误差和信噪比为评价指标,对信号采集系统获取的煤与瓦斯突出AE信号分别使用标准粒子滤波、果蝇优化粒子滤波、改进粒子滤波去噪。结果表明,改进粒子滤波法的信噪比提升了15. 3 d B左右,且均方根误差最低。与其他两种方法相比,改进粒子滤波去噪效果最优。更多还原

【Abstract】 Coal-gas outbursts produce acoustic emission signals( AE). Giving the problem of extracting more pure and effective AE signals,this paper proposed an intelligent optimized particle filter( PF) of neighborhood dynamic adjustment( D) for fruit fly algorithm( FOA). It used each fruit fly individual characterizing signal particle of PF to optimize the resampling process of particle filtering,and adjusting the number of neighboring particles dynamically improved the optimization ability and convergence rate of fruit fly algorithm. Taking the root-mean-square error and signal-to-noise ratio as the evaluation index,from signal acquisition system,it respectively de-noised the AE signals of coal-gas outburst by standard particle filter,fruit fly optimized particle filter and improved particle filter. The experiment results show that the signal-to-noise ratio of the optimized particle filter algorithm is improved by about 15. 3 d B,and the root-mean-square error is the lowest. Compared with other two methods,the improved particle filter has the best de-noising effect.更多还原