基于自适应移位平均降噪与BP神经网络的钢丝绳损伤识别

Damage identification of the mining wire rope based on adaptive moving average denoise and BP neural network

提出了一种基于自适应移位平均降噪与BP神经网络的钢丝绳损伤识别方法,解决了强噪声背景下的钢丝绳损伤识别问题。以矿井钢丝绳为检测对象,采用自适应移位平均法对含噪的断丝信号与磨损信号进行降噪处理,通过自适应粒子群优化(APSO)算法找到移位平均算法的最优窗宽;然后,以断丝损伤为例,对输出的最优降噪信号提取峰峰值、波宽、波形下面积三种特征值作为特征值样本,将样本归一化后输入BP神经网络进行损伤识别;最后,通过试验验证了所提方法的有效性。试验结果表明,该方法能定性识别钢丝绳损伤并且识别准确率高。

A wire rope damage identification method based on adaptive displacement average noise reduction and BP neural network was proposed to solve the problem of wire rope damage recognition under the background of strong noise.Taking the mining wire rope as an example, firstly, the adaptive moving average method was used to denoise the broken wire signal and wear signal, and the optimal window width of the moving average algorithm was found through the adaptive particle swarm optimization(APSO);Then, taking the broken wire damage as an example, three eigenvalues of the output optimal noise reduction signal were extracted as eigenvalue samples: peak to peak value, wave width and area under waveform. The samples were normalized and put into BP neural network for damage identification;Finally, the effectiveness of the proposed method was verified by experiments. The experimental results show that this method can qualitatively identify the wire rope damage, and the recognition accuracy is high.