【作者】 成玫；

【导师】 胡雄；

【作者基本信息】 上海海事大学 ， 机械电子工程， 2004， 硕士

【摘要】 随着我国加入WTO，我国港口在进出口贸易中的地位更加显得突出。90年代以来，是世界集装箱运输飞速发展的时期，与之相对应的是集装箱桥吊(以下简称岸桥)的大型化和高效化。在岸桥前大梁与后大梁处有一连接铰点，每次作业的循环中，小车都要负载或空载通过该铰点，这会对铰点处的轨道接头产生一定的冲击和振动，这些冲击和振动会带米许多危害，若不及时发现，将会对整机的工作性能、效率产生重要的影响。 近二十多年来，我国一些高等院校、研究所和工厂，在振动监测和故障诊断这一领域开展了许多研究工作，并取得了不少成果，但是对于岸桥轨道铰接处振动的监测和故障诊断则不够完善。本论文即是以此为研究的切入点，通过在岸桥模型上进行模拟试验，收集岸桥铰接处轨道在不同高低差、不同的车速和载重下的信号，应用小波包分解的统计量的分析技术对这些数据进行分析。 在应用小波包分解时，存在着小波基的选择问题，不同的小波基具有不同的时频特性，因此同一信号经过不同小波基进行分析往往会表现出不同的结果。本文分析了常用一维小波基的特性，针对本试验的特征，比较了Haar小波基、Daubechies小波基、Coiflet小波基、Symlets小波基、Bior小波基、Meyer小波基、Morlet小波基、Mexican Hat小波基，在这几种小波基中，找出了最适合本试验的小波基——Bior3.3。 统计量的分析方法多种多样，本文选用了峭度指标、裕度指标分析小波包分解后各频段的数据，经过大量数据的比较，发现选用峭度指标、裕度指标对数据分析有帮助：由峭度指标能区别出两侧轨道中是哪一侧出故障；应用裕度指标找出了在岸桥铰接处轨道的不同高度差下，该指标的变化规律。同时在计算峭度指标、裕度指标中，通过比较认为平滑法计算较精确，并选用了该方法计算统计量。 本论文由测试参数的确定、测试点的选择、测试试验硬件的配置及测试工况的确定，拟定了在岸桥模型上模拟岸桥主铰处故障的测试试验方案。通过对测试试验数据的小波包分解的统计量的分析，发现经过尺度为5的小波包分解后，在2625-2750Hz频段内，统计量——平滑裕度指标随着岸桥铰接处轨道的高度差的增加而增加。由于振动数据是在岸桥模型上模拟而获得的，具有一定的可信度，因此该分析结果对以后的实际应用有一定的指导作用!更多还原

【Abstract】 The port of our country is more outstanding in the position of import-export trade as our country joins the World Trade Organization. Since the 1990s, it was the period that the containerized transport of the world developed at full speed, what the ones that corresponded to it in the quayside container crane (crane for short). And the crane is more large-scaly and high-efficiently. There is a hinge between the summer beam. The trolley always transits the hinge whenever it loads or unloads. It will cause impaction and vibration. Then these impaction and vibration will make important influence in the efficiency working of crane, if we don’t discover those in time.In the last 20 years, there were lots of colleges, institutes and factories studied in the vibration detection and failure diagnosis. As a result, it has developed a lot of research work in this field. But the vibration detection and failure diagnosis to the hinge of crane is not perfect. Thus this thesis investigates in the point. Though the imitating experiment in crane model, there are a great deal of different signals are collected in different cases. Then the analysis technique of the Wavelet Packets and Statistic are applied.In using the Wavelet Packers, there is an issue in choosing the wave function. Because different wave functions has different time domain specification and frequency characteristic. It will express a different result that different wave functions are applied in the same signal. In this thesis, one dimension wave functions in common use is analyzed. Compared with these wave functions of Haar, Daubechies, Coiflet, Symlets, Bior, Meyer, Morlet and Mexican Hat, the wave function of Bior3.3 is picked out considering the character of this experiment.There are various methods in the analysis of statistics. Kurtosis Value and Clearance Factor were used to analysis these signals that have been decomposed by the Wavelet Packets. And Kurtosis Value and Clearance Factor were good after a good deal of data was compared. By Kurtosis Value, it can distinguish out which side has fault. And a law can be discovered by Clearance Factor. By virtue of the accuracy of the smoothing algorithm, it is used in the calculation of statistics.In this thesis, though establishing test parameters, test points, test hardwares and test cases, test method plan to analog the fault in the crane hinge by crane model. After dimensions for 5 of the Wavelet Packets and statistic analysis,statistics-Smooth Clearance Factor are increased as the increase of thedifference in height of the crane hinge, in 2625-2750 Hz frequency channels.These vibration data are credible because those collected from the simulation in the crane model. Therefore the analysis has certain guidance functions to the practical application for the future in the result!Cheng Mei (Mechanical electronic engineering) Directed by: Prof.Hu Xiong更多还原