首页|毛纺机车间智能监测与故障诊断技术研究

毛纺机车间智能监测与故障诊断技术研究

扫码查看
为了提高毛纺机运行稳定性、降低故障停机时间以及优化企业生产流程,改善传统毛纺机车间智能监测与故障诊断方法对于潜在故障的识别效果较差的问题,设计新的毛纺机车间智能监测与故障诊断技术.通过振动传感器、红外传感器、位移传感器、图像传感器对毛纺机车间设备进行信号采集;设计由服务器、远程监测客户端、数据分析客户端组成的智能远程监测装置,各部分通过以太网相互连接,用于实施毛纺机车间的智能监测;设计融合VAE模型与 LSTM 模型优点的 LSTM-VAE混合模型,将传感器采集数据作为输入,通过自适应阈值算法为车间设备自适应地计算出预警阈值,实现毛纺机车间的故障诊断.测试结果表明,设计技术判定的潜在故障点与实际的潜在故障点十分接近,通过设定自适应阈值,使得该技术的诊断结果与设备自身情况更加贴合.
Research on intelligent monitoring and fault diagnosis technology for woolen spinning machine workshop
In order to improve the running stability of woolen spinning machine,minimize the downtime of faults and optimize the production process of enterprises,a new technology of woolen spinning machine workshop intelligent monitoring and fault diagnosis was designed to improve the effectiveness of traditional intelligent monitoring and fault diagnosis methods in identifying potential faults.Vibration sensor,infrared sensor,displacement sensor and image sensor were used to collect signals from woolen spinning machine equipment.An intelligent remote monitoring device composed of server,remote monitoring client and data analysis client was designed.All parts were connected by Ethernet to implement intelligent monitoring among woolen spinning machine workshop.The LSTM-VAE hybrid model was designed,which combines the advantages of VAE model and LSTM model.The sensor data was used as input,and the early warning threshold was calculated for the workshop equipment through adaptive threshold algorithm,so as to realize the fault diagnosis among woolen spinning machine workshop.The test results show that the potential fault points determined by the design technology are very close to the actual potential fault points.By setting the adaptive threshold,the diagnosis results of the technology are more consistent with the equipment itself.

sensorsLSTM-VAE hybrid modelwoolen spinning machine workshopstatus monitoringfault diagnosis

吴文贤、沈博侃

展开 >

浙江纺织服装职业技术学院 机电与轨道交通学院,浙江 宁波 315211

传感器 LSTM-VAE混合模型 毛纺机车间 状态监测 故障诊断

2024

毛纺科技
中国纺织信息中心 北京毛纺织科学研究所

毛纺科技

北大核心
影响因子:0.3
ISSN:1003-1456
年,卷(期):2024.52(11)