基于拍摄装置自动调节的TFDS设备故障识别  

Fault Identification of Tfds Equipment Based on Automatic Adjustment of Shooting Device

作　　者：丁凤霞 刘振中 DING Feng-xia;LIU Zhen-zhong(CHN Energy Investment Group Co.,Ltd.,Beijing 100120 China;Beijing Jingtianwei Technology Development Co.,Ltd.,Beijing 100085 China)

机构地区：[1]国家能源投资集团有限责任公司,北京100120 [2]北京京天威科技发展有限公司,北京100085

出　　处：《自动化技术与应用》2023年第12期71-74,89,共5页Techniques of Automation and Applications

摘　　要：为了获得理想的故障识别结果,提出基于拍摄装置自动调节的TFDS设备故障识别方法。该方法首先自动拍摄货车故障图像,采用均值滤波器去噪处理拍摄的故障图像,并将处理完成图像分成训练和测试样本,训练样本经卷积和池化操作后,通过随机梯度下降算法更新参数构建卷积神经网络模型,经测试样本测试后,使用softmax回归函数输出货车故障识别结果。实验证明TFDS设备故障识别效果好,方法具有较高的实际应用价值。In order to obtain ideal fault identification results,a fault identification method of TFDS equipment based on automatic adjust-ment of shooting device is proposed.Firstly,the fault image of freight train is automatically taken,the fault image is denoised by mean filter,and the processed image is divided into training and test samples.After convolution and pooling,the parameters of the training samples are updated by random gradient descent algorithm to construct the convolution neural network model.After testing the test samples.The softmax regression function is used to output the fault identification results of freight cars.Experi-ments show that the TFDS equipment fault identification effect is good and this method has high practical application value.

关 键 词：自动调节 TFDS设备 PLC控制 卷积神经网络 滚珠丝杠 均值滤波器 

分 类 号：TP206.3[自动化与计算机技术—检测技术与自动化装置] TP273[自动化与计算机技术—控制科学与工程]