铝合金GTAW焊接气孔缺陷的在线预测与实时控制是焊接领域最令人困惑的难题之一。本项目试图探索一种基于电弧信息特征智能处理的铝合金GTAW焊接气孔缺陷的在线预测及实时抑制智能控制的新途径。主要内容包括：探索铝合金焊接气孔形成特征的可检测机理和可控制机理，尝试发展一种系统性方法：基于铝合金GTAW焊接电弧光谱特征检测获取电弧电子温度Te曲线；探究其与铝合金GTAW焊接气孔缺陷形成的规律；结合铝合金GTAW焊接动态过程光谱、熔池图像、电弧声音等信息以及焊接工艺多参数影响因素，给出铝合金GTAW多传感信息特征智能分类与融合计算方法；建立铝合金GTAW焊接气孔缺陷形成预测的粗糙集知识模型；探索发展合适于铝合金GTAW焊接气孔缺陷形成过程抑制的智能控制方法，为解决铝合金GTAW焊接气孔缺陷在线预测与实时抑制控制的难题提供有价值的科学方法和技术实现途径。

The on-line prediction and prevention of porosity defects during welding process is one of the most perplexing problems in welding field. In this proposal, we will propose a new way of on-line prediction and real-time prevention of typical porosity defects in Al alloy GTAW based on intelligent processing of arc characteristic information . The main content includes as following: investgate the measurable,and controllable mechanism of typical porosity defects during Al alloy GTAW; develop a series of methods：take the Te curves of arc electron temperature based on the characteristic of arc and spectrum in Al alloy GTAW process; investigate the characters of Te curve and formation principles of typical porosity defects based on the analysis of arc temperature and mechanism of arc continuous spectrum during Al alloy GTAW; combine with the spectrum, pool image arc audio and multi-information of dynamic welding process and regard of the influence of multi-parameters in Al alloy GTAW process, based on intelligent classification of multi-sensor information and fusion calculation, rough set knowledge extraction and model, then explore to develop some new intelligent control methods to restrain and prevent the typical porosity defect formation during Al alloy GTAW process. The research will provides a scientific method and technical approach to achieve some significant results on the on-line prediction and real-time prevention of typical porosity defects for Al alloy GTAW.