基于机器视觉的电机故障识别方法及实验研究

针对信号处理故障检测方法难以自动辨识故障类型的问题,提出了一种基于机器视觉的感应电动机故障识别方法。首先利用Park矢量法分析电机的三相电流,获取矢量轨迹图;然后将方向梯度直方图与支持向量机相结合实现图像的特征提取和分类,进而达到电机故障自动辨识的目的。在Maxwell和Matlab的仿真环境下,设计了多种故障电机的模型和相应的仿真实验方案,实验结果表明:该文所提方法不但可行,同时其故障识别率高达100%。在此基础上,搭建了实物实验平台,进一步验证了所提方法对电机故障的高识别率。整个方法的研究和验证过程在教学中可加深学生对电机原理、信号处理以及人工智能等相关知识的理解,提高他们的学习兴趣和理论联系实际的能力。

Motor fault identification method based on machine vision and its experimental research

In view of the problem that the fault type cannot be automatically identified in the signal processing fault detection method, a fault identification method based on machine vision for induction motors is proposed. Firstly, the Park vector method is used to analyze the three-phase current of the motor to obtain the vector track maps. Then the histogram of oriented gradient is combined with the support vector machine to realize the extraction and classification of image features so as to achieve the purpose of automatic identification of motor faults. In the simulation environment of Maxwell and Matlab, a variety of fault motor models and corresponding simulation experiments are designed. The experimental results show that the proposed method is not only feasible, but also has a fault recognition accuracy of 100%. On this basis, a physical experiment platform is built to further verify the high recognition rate of the proposed method for motor faults. The whole process of research and verification of the method can deepen students’ understanding of motor principle, signal processing, artificial intelligence and other relevant knowledge in teaching, and improve their learning interest and ability to combine theory with practice.