【作者】 何超；

【导师】 廖义德；

【作者基本信息】 武汉工程大学 ， 机械电子工程， 2017， 硕士

【摘要】 随着汽车工业的高速发展,汽车发动机的性能不断地改善和提高。轮系作为发动机的必备配件,其质量直接决定了发动机的工作效率。我国在发动机轮系配套设施研究方面尚属起步阶段,轮系设计周期长、实验成本较高,并且需要反复的修订,直接导致国内自主汽车配套设施设计成本的增加。汽车发动机轮系的复杂性主要表现在工作时产生的各类振动,为了减少系统振动,增加乘用人员的舒适性,需要对轮系的相关理论和分析方法展开更加深入的研究。对此,本文主要开展了以下研究工作:1.以发动机轮系多楔带为研究对象,建立了多楔带动态特性采集平台。首先利用摄像机采集轮系时序图像信息,并采用图像处理对轮系标示点进行图像提取、灰度变换、颜色特征提取、边缘检测等从而获得多楔带标示点单个像素的精确坐标,然后跟踪轮系运动过程中多帧图像的同一个标示点,得到轮系连续运动轨迹。2.为解决摄像机采集轮系数据出现的噪点、拖影与畸变而引起的数据丢失、误差等问题,采用等距抽样方法提取多楔带弦线节点数据并运用5种不同的插值方法对振动数据进行重构,将摄像机实测到的全部振动数据与分析结果进行对比,将每个插值算法的结果误差进行分析,得出误差最小的插值算法,弥补了多楔带数据信息缺失的问题。3.建立了轮系动态特性的数学模型。首先将轮系的各带段简化成轴向运动弦线,应用伽辽金法将带的时间-空间连续方程离散成为时间和空间函数之积并且对发动机轮系动态特性进行分析。然后研究了多楔带的波动形状、横向振动位移、振幅、频率以及从动轮-带滑移率等参数。最后将实验结果与计算结果进行对比分析,验证了测试方法和计算模型的正确性,从而完善了轮系动态特性分析方法的研究。经过实验研究,本文所提出的基于时序图像处理的汽车发动机轮系动态特性分析方法,在实际动态特性分析中的效果较好,能够满足轮系振动检测的准确性,可为轮系的设计和发展提供参考依据。更多还原

【Abstract】 With rapid development of the automobile industry,performance of Automobile engine continues to be improved and enhanced.The quality of gear train,which is regarded as an essential part of the engine,is significant for the efficiency of the engine.And the research of the engine gear train facilities in China is still in the initial stage,the deficiency namely long cycle design,higher experimental cost,and repeated amendments lead to an increasing cost of the domestic automobile industry.Additionally,the complexity of the automotive engine gear train cause a variety of harmful vibration.In order to reduce system vibration and increase the comfort,more in-depth research of the theory and algorithm of gear train need to be carried out.In this paper,the following research work has been carried out:1.The dynamic characteristics acquisition platform of serpentine belt is established to conduct the research about the serpentine belt of engine.Firstly,the timing image information of gear train is acquired by a camera.Then for the purpose of obtaining the exact coordinates of the individual pixels of the Serpentine belt mark,image processing techniques are used to extract the image,transform gray scale extract color feature and detect the edge.Finally the same mark point of the multi-frame image is tracked during the movement of the gear train so that the continuous trajectory of it is collected.2.In order to solve the data loss and error issues caused by camera,equidistant sampling theorem is applied to extract the data of the Serpentine belt string nodes.Moreover,five different interpolation approaches are adopted to reconstruct the data.By comparing the results of interpolation approaches with the data measured by the camera,error analysis is designed to calculate the interpolation algorithm with the least error and enhance the data quality.3.A mathematical model of gear train dynamic characteristics is developed in the study.Firstly,the belt sections of gear train are simplified into axial motion strings.Then Galerkin method is applied to discrete the time-space continuum equations of belt into the product of time and space functions and calculate the dynamic characteristics of theengine gear train.Additionally,parameters such as fluctuating shape,lateral vibration displacement,amplitude,frequency and follower wheel-slip ratio of the Serpentine belt are studied.Finally,for the purpose of improving the study on the dynamic characteristics analysis method of the gear train,experimental results are compared with calculation to verify the accuracy of model and test method.In this paper,the dynamic characteristic analysis method based on the time series image processing is proposed and proved to be an effective method in the analysis of the dynamic characteristics in practical.And it can meet the requirement of the wheel vibration detection and provide references for the development of gear train.更多还原