This study focuses on enhancing weld joints of AA2017 aluminum alloys obtained through the UFSW (Underwater Friction Stir Welding) process. An integrated method utilizing Grey relational entropy measurement as a multi-response optimization tool was employed. The objective was to improve the performance of the weld joint by optimizing the process parameters, including advanced speed, tool rotation speed, and forging force, to enhance tensile strength, elongation, and microhardness. The L9 orthogonal design with three factors and three levels was utilized, and subsequent analysis of variance and confirmation tests were conducted. Furthermore, the influence of lateral deviation of the welding tool on the mechanical properties and microstructural characteristics of the optimal joint was investigated. Experimental results aligned well with the outcomes predicted by the proposed mathematical model, affirming the effectiveness of the optimization method. Notably, the study demonstrated an increase in weld joint efficiency from 68.55% to 93% through the implementation of the optimization approach. The welding tool deviation ranging from -1.5 mm to 1.5 mm showed that the tensile strength of the optimal joint decreases while the strain rate increases when the offset was on the advancing side (AS).