高速重载机车牵引系统多元故障劣化趋势预警方法研究

Modern high-speed and heavy-haul locomotive traction system is an intelligent system combination of Mechanical, electronic and information technology. Once faulty happen power performance and reliability decreased rapidly, which directly affect safety and stability of whole system. Take the load spectrum of traction system under many fault conditions for foundation in this project, which reveal fault trend development characteristics of stable operation deterioration. Building multi-body dynamics model based on stability theory binding ANSYS. Analyze the effect of multi-parameters on the fault deterioration trend, integrated sensors diversity wireless detection model was built, which realized remote seamless information transmission, and anti-interference capability was enhanced. Based on the dynamic sample information, multi-fault data association fusion diagnosis and efficient acquisition fault characteristics of poor information were researched in order to build fault deterioration knowledge base, which can improve the reliability and accuracy of remote fault forecast on high-speed and heavy-haul locomotive.

现代高速重载机车的牵引系统是集机械、电子、信息于一体的智能化系统，关键功能部件一旦发生故障，其动力性和可靠性将迅速下降，直接影响着整个系统的安全性和稳定性。本项目拟以多故障工况下牵引系统的载荷谱为基础，揭示由稳定运行状态劣化为非稳定运行状态的故障趋势发展特征，结合有限元分析建立基于稳定性理论的多体动力学模型。通过分析多检测参数对故障劣化趋势的影响，建立集成传感器多元化无线检测模型，实现远程无缝信息传输，提高抗干扰能力。以多动态样本信息为基础，研究多故障数据关联融合诊断，有效获取微弱信息故障特征，建立故障劣化知识库，对提高高速重载机车远程故障预警的可靠性和准确性具有重要意义。

以现代机车牵引系统的关键功能部件的可靠性和稳定性为研究对象，开展了齿轮啮合全生命周期的图像检测方法研究，实现了齿轮啮合面在不同啮合力状态下的检测，为建立齿轮啮合动力学模型提供了分析基础；研究了齿轮的动态检测方法，实现了基于时间的动态检测，为揭示齿轮传动系统故障劣化机理研究提供了关键条件；建立了基于齿轮传动失效机理齿面胶合动力学模型，为提高稳定性区域和传动系统机械动态特性提供了优化方法；研究了一种自适应时变稀疏盲分离方法，有效提高了故障特征识别准确率，避免了信息量缺少、噪声大等不稳定状况的发生；针对机车齿轮箱多信号受到不确定性因素影响而存在微弱性和混叠性难以识别问题，研究了一种动态级联经验模态分解方法，能够有效抑制模态混叠现象并且实现有效获取故障特征频率，实现了在大量噪声背景下的微弱故障特征获取问题，为及时、准确进行早期故障预示及诊断提供了一种有效手段。本项目研究内容对提高机车远程故障预警的可靠性和准确性具有重要的实际意义。