2D LiDAR simultaneous localization and mapping(SLAM) often encounters issues such as registration failures, cumulative errors, and low scan repeatability, which prevent the construction of consistent and accurate maps. To address these challenges and reduce the need for remapping, this study proposes an interactive solution for 2D LiDAR SLAM, commonly used in AGVs, by manually adding or modifying constraints. However, for large-scale environments, manually adding and correcting constraints can be highly labor-intensive and difficult. To overcome this, an automatic loop closure detection and edge enhancement strategy was developed. A multi-scale loop closure detection algorithm enables fast loop closure detection, while the edge enhancement strategy revisits the pose graph to reduce overall error. Additionally, a GUI application was developed to visualize the pose graph, edges, and node information, allowing for interactive batch optimization. Finally, several experiments were conducted to demonstrate the performance improvements achieved by the proposed method.