Intelligent computing technology is driving progress in economy and society while fostering deeper innovations in scientific research. As an interdisciplinary field that studies the complex large-scale system of human-water interactions, the integration of intelligent computing technology enhances research methodologies and promotes the development of the human-water relationship discipline. Therefore, this paper aimed to propose and establish an intelligent computing framework for human-water relationship discipline and explore its application potential through application analysis. The proposed framework consists of four key components: intelligent identification, intelligent assessment, intelligent simulation, and intelligent optimization and regulation. Based on CNN-LSTM-Attention model combined with watershed supply-demand balance calculations, the feasibility and effectiveness of intelligent computing in simulating and optimizing human-water relationships were examined through case application. The results indicate that the intelligent computing framework holds broad application prospects in five key areas: human-water interaction mechanisms, dynamic processes, simulation and prediction, scientific regulation, and policy formulation. The CNN-LSTM-Attention model demonstrated high accuracy in runoff simulation for the Qin River Basin, with training and simulation accuracy parameter achieving R2 values reaching 0.93 and 0.88, respectively. Furthermore, the supply-demand balance index for the 2030 planning horizon is 1.34, indicating a tight water supply-demand relationship. This study provides new perspectives and insights for advancing human-water relationship discipline and improving regional water resource management.