Abstract: With the development of offshore equipment towards deep water and large scale, applications of high-performance thick plates are more and more extensive, and welding quality requirements of thick walled structural parts sensitive to heat input are gradually increasing. Narrow-gap laser welding can reduce deformation and residual stress of welded joints while reducing filling amount of weld and improving production efficiency. In order to control quality of narrow-gap welding process, it is necessary to develop an accurate and effective seam tracking and welding process monitoring system. The signal acquired by a single sensor is limited, and it is difficult to fully reflect welding process information. Multi-sensor-based information fusion technology is applied to monitoring laser narrow-gap welding process. By using multi-source sensing technology and selecting a proper information fusion algorithm, a multi-source information fusion model suitable for the narrow gap laser welding process can be established to obtain more accurate and comprehensive information describing state of welding process, and precision and efficiency of seam tracking are expected to be improved. It has important scientific significance and economic value for realizing high-quality and efficient welding of thick-walled structural parts and promoting development and progress of marine equipment manufacturing technology in China.