基于神经网络技术的机器人并发故障自动诊断方法

为了提高柔性负载抓握机器人的故障检测能力,提出基于神经网络技术的机器人并发故障自动诊断方法。运用高分辨的智能传感器信息识别技术,结合刚度和强度等机械结构特征分析,构建柔性负载抓握机器人的故障信息采集模型,采用变刚度原理,提取柔性负载抓握机器人的振荡信息特征,通过谱特征检测和动态信息融合进行柔性负载抓握机器人的故障信息的多分辨融合和特征聚类处理。通过分析故障样本信息数据参数的估计值,对信息数据进行重组,根据采样信息的差异性对故障类别进行初步判断和识别。采用BP神经网络技术,通过特征分布函数进行故障特征提取,进行机器人并发故障的优化诊断和自适应学习,提高机器人并发故障的有效检测和识别能力。仿真结果表明,采用该方法进行机器人并发故障诊断的自适应性较好,特征辨识能力较强,具有很好的故障监测和模式识别能力。

Automatic fault diagnosis method for robot based on neural network technology

In order to improve the fault detection ability of flexible load grasping robot,an automatic fault diagnosis method based on neural network technology is proposed.Using high-resolution intelligent sensor information recognition technology,combined with the analysis of mechanical structure characteristics such as stiffness and strength,the fault information collection model of the flexible load grasping robot is constructed.The principle of variable stiffness is used to extract the vibration information characteristics of the flexible load grasping robot.Feature detection and dynamic information fusion are used for multi-resolution fusion and feature clustering of the fault information of the flexible load gripping robot.By analyzing the estimated values of the data parameters of the fault sample information,the information data is reorganized,and the fault category is preliminarily judged and identified based on the difference of the sampled information.Using BP neural network technology,fault feature extraction through feature distribution function,optimized diagnosis and adaptive learning of robot concurrent faults,and improved effective detection and recognition capabilities of robot concurrent faults.Simulation results show that this method has good adaptability,strong feature recognition ability,and good fault monitoring and pattern recognition ability.