【作者】 白凤民；

【导师】 单平；

【作者基本信息】 天津大学 ， 材料加工工程， 2004， 硕士

【摘要】 电阻点焊质量监测技术是一种高效率、低成本的质量保障措施。然而，目前的监测技术主要是根据点焊过程参数的某一特征量与质量参数之间的线性回归模型提供的信息来间接的监测焊点质量。由于实际点焊过程的复杂程度使这种模型与实际情况相离较远，只能在特定的条件内提供点焊的质量信息，致使这种监测技术存在监测效果不够稳定、适用范围窄等问题。为了解决这些问题，就需要充分利用监测信息，详细研究监测信息与质量参数之间的函数关系，使建立的模型能够在较宽的条件内提供准确、可靠的点焊质量信息。铝合金点焊在汽车工业中的应用越来越普遍，为了解决铝合金点焊的质量监测问题，本文开发了基于多信息融合的铝合金直流点焊质量监测系统。该系统采用 Visual C++与 MATLAB 相结合作为系统编程语言；利用研制的检测系统，对点焊过程中电流信号、电压信号、位移信号和声音信号等反映点焊质量的物理量进行同步数据采集，并将采集到的数据经数据预处理模块提取特征值之后送入融合中心，用 BP神经网络方法进行多信息融合后，得出熔核的面积，并根据面积值来判断焊点质量是否合格。试验结果表明，该系统能够正确的采集焊接过程中的焊接电流、焊接电压、电极位移及声音等信号；采用合理的信号处理方法，将各信号的特征值进行多信息融合，其结果可以用来判断焊点的焊接质量。更多还原

【Abstract】 Monitoring and control technique for resistance spot weldingquality is a high efficiency, low cost quality assurance measure. However,the techniques developed are to monitor spot welding quality indirectlyby means of quality information, which is provided in welding process bythe regression models between quality parameter and a certain feature ofprocess parameter. The actual spot welding process is so complicated thatthe models are far away from the actual object, the models can onlysupply reliable information about weld quality within narrow conditions,which causes control effect unstable and applicability narrow, etc. To solve the above problems, the monitoring information must bestylized fully, and the model relations between monitoring informationand quality factors must be studied in detail, so that the models cansupply accurate and reliable spot welding quality information within wideconditions. Aluminum alloys resistance spot welding is widely used inautomobile industry. In order to solve the quality monitoring of aluminumalloys resistance spot welding, quality monitoring system based onmulti-information fusion for aluminum alloys DC spot welding wasdesigned in this paper. They system was design by Visual C++ andMATLAB, using simultaneous data collection card for data collection.The data collected were treated by data processing module and then fusedby BP ANN in fusion center. Nugget area was gained to estimate thequality of welding spots. The results suggest that the system has uniform sampling intervaland can collect signals such as welding current, welding voltage,electrode displacement, sound signals etc. The relationship betweenwelding condition and welding quality can be gained by fusingeigenvalues of signals.更多还原