基于周期势函数的自适应二阶欠阻尼随机共振信号增强方法

针对滚动轴承微弱故障振动信号在噪声环境下故障特征难以提取的问题,提出一种基于周期势函数的自适应二阶欠阻尼随机共振信号增强方法。采用粒子群算法对系统参数和阻尼系数的自适应匹配,实现对多个拟增强频段的随机共振,更加适用于工程实际中多故障信号提取。数据库考题检验和工程实验验证表明:1)该方法明显提高了输出信噪比,故障特征频率处主峰突出,边带干扰少,方便故障的机器判读,误判率低;2)随着噪声强度的增加,虽然输出信噪比有所降低,但该方法的检测效果仍优于基于周期势函数的自适应一阶随机共振方法的检测效果;3)该方法对噪声的适应性更强,在噪声环境下对于微弱故障信号的提取有着明显优势。

Adaptive Second-order Underdamped Stochastic Resonance Signal Enhancement Method Based on Periodic Potential Function

Aiming at the problem that the faulty vibration signal of rolling bearing is difficult to extract under noisy environment, an adaptive second-order underdamped stochastic resonance signal enhancement method based on periodic potential function is proposed. The method uses particle swarm optimization to measure system parameters and damping. The adaptive matching of the coefficients realizes the stochastic resonance of multiple pseudo-enhanced frequency bands, and is more suitable for multi-fault signal extraction in engineering practice. The test of fault signal and the engineering experiment show that: 1) The adaptive second-order under-damped stochastic resonance method based on periodic potential function significantly improves the output signal-to-noise ratio, the main peak of the fault characteristic frequency is prominent. The sideband interference is less, the machine is easy to fault, and the false positive rate is low. 2) With the increase of noise intensity, although the output signal-to-noise ratio is reduced, the adaptive second-order underdamped stochastic resonance method based on periodic potential function, the detection effect is still better than the adaptive first-order stochastic resonance method based on periodic potential function. 3) The adaptive second-order under-damped stochastic resonance method based on periodic potential function is more adaptable to noise, in noisy environment, there are obvious advantages for the extraction of weak fault signals.