| 融合SVM和传感器数据的自动扶梯踏板振动故障检测方法 |
 点此下载全文 |
| 引用本文：李自翔.融合SVM和传感器数据的自动扶梯踏板振动故障检测方法[J].计算技术与自动化,2025,(2):55-59 |
| 摘要点击次数: 924 |
| 全文下载次数: 766 |
|
|
| 中文摘要:为了实现自动扶梯踏板振动故障精准检测，提出了融合SVM和传感器数据的故障检测方法。使用自动扶梯踏板振动故障传感器，结合FFT频谱分析方法构建矩阵，对倒频谱平滑处理。模拟平滑处理过程，建立自回归滑动平均模型，剔除光滑信号和噪声后的突变信号。构建故障特征融合SVM模型，运用拉格朗日函数求解最优超平面的问题。设计基于SVM故障检测步骤，设定最优分类阈值，完成自动扶梯踏板振动的故障检测。由实验结果可知，所研究方法X、Y、Z方向振动故障信号波动幅值分别为\[-0.02 dB，0.02 dB\]、\[-0.01 dB，0.02 dB\]和\[-0.06 dB，0.05 dB\]，只在Z方向与实际波动幅值存在0.01 dB的幅值波动误差，其余均一致，具有精准检测效果。 |
| 中文关键词:融合SVM 传感器数据 自动扶梯 踏板振动 故障检测 |
| |
| An Automatic Escalator Pedal Vibration Fault Detection Method Based on SVM and Sensor Data Fusion |
|
|
| Abstract:In order to achieve accurate detection of escalator pedal vibration faults, a fault detection method integrating SVM and sensor data is proposed. Using an escalator pedal vibration fault sensor and combining FFT spectrum analysis method to construct a matrix for smooth processing of the cepstrum. Simulate the smoothing process, establish an autoregressive moving average model, and eliminate abrupt signals after smoothing and noise. Construct a fault feature fusion SVM model and use the Lagrange function to solve the problem of optimal hyperplane. Design fault detection steps based on SVM, set the optimal classification threshold, and complete fault detection of escalator pedal vibration. According to the experimental results, the fluctuation amplitudes of the vibration fault signal in the X, Y, and Z directions of the studied method are \[-0.02 dB，0.02 dB\], \[-0.01 dB，0.02 dB\], and \[-0.06 dB，0.05 dB\], respectively. There is only a 0.01 dB amplitude fluctuation error between the Z direction and the actual fluctuation amplitude, and the rest are consistent, indicating accurate detection effect. |
| keywords:fused SVM sensor data escalator pedal vibration fault detection |
| 查看全文 查看/发表评论 下载pdf阅读器 |
|
|
|
位访问者,今日访问
