Against the backdrop of sustainable development of green cities, eco-environmental security has become increasingly critical, with real-time monitoring and prediction of deformation in water conservancy dams under complex environments being of paramount importance. This study introduces visual design based on the management of original dam monitoring data to achieve real-time, dynamic, and multi-dimensional information display through line charts. AnImproved Fruit Fly Optimization Algorithm (IFOA) IFOA-BP-AdaBoostIFOA-BP-AdaBoost strong prediction model is constructed, and comparative experiments on prediction accuracy are conducted with BP andImproved Fruit Fly Optimization Algorithm (IFOA) IFOA-BP algorithm models. Taking monitoring data from a specific dam as a case study, the results show that: (1) Incorporating visual design into the prediction process enhances the readability and comprehensibility of data information; (2) The root mean square error (RMSE) of theImproved Fruit Fly Optimization Algorithm (IFOA) IFOA-BP-AdaBoostIFOA-BP-AdaBoost model is 0.3597 mm, demonstrating accuracy improvements of 33.67 and 11.19% compared to the BP andImproved Fruit Fly Optimization Algorithm (IFOA) IFOA-BP models, respectively. The proposed methods are expected to provide technical support for dam monitoring and prediction in complex environments.