论文部分内容阅读
为提高双级滑阀真空泵工作可靠性,针对双级滑阀式机械真空泵在实际生产过程中可能出现的故障现象,引入控制律重组和PARD-BP神经网络故障诊断算法,以2H-150型双级滑阀真空泵振动故障样本采集数据作为输入,以故障模式矩阵作为目标输出,对PARD-BP算法所得数据进行训练优化;再提出动态数据信息提取概念,进行其动态信息收集提取,采用目前广泛应用的Poly-Max模态参数识别方法进行结果验证,结果表明真空泵X、Y和Z方向上最大的峰值分别下降了31.93%,21.54%和19.37%,证实滑阀真空泵的故障诊断方法具有可靠性。
In order to improve the working reliability of vacuum pump with double-stage slide valve, aiming at the fault phenomenon that double-stage slide valve vacuum pump may have in actual production process, the control law recombination and PARD-BP neural network fault diagnosis algorithm are introduced. Level pump vacuum pump vibration fault sample data as input, the fault pattern matrix as the target output, the PARD-BP algorithm for training data to optimize; then put forward the concept of dynamic data extraction, dynamic information collection and extraction, using the widely used The results show that the maximum peak value of vacuum pump in X, Y and Z directions is decreased by 31.93%, 21.54% and 19.37% respectively, which proves the reliability of the fault diagnosis method of slide valve vacuum pump.