Digital twins (DTs) have seen widespread application across industries, enabling deep integration of cyber–physical systems. However, previous research has largely focused on domain-specific DTs and lacks a universal, cross-industry modeling framework, resulting in high development costs and low reusability. To address these challenges, this study proposes a DT modeling method based on hierarchical decoupling and topological connections. First, the system is decomposed top–down into three levels—system, subsystem, and component—through hierarchical functional decoupling, reducing system complexity and supporting independent component development. Second, a method for constructing component-level DTs using standardized information sets is introduced, employing the JSON-LD language to uniformly describe and encapsulate component information. Finally, a topological connection mechanism abstracts the relationships between components into an adjacency matrix and assembles components and subsystems bottom–up using graph theory, ultimately forming the system-level DT. The effectiveness of the proposed method was validated using a typical surface water purification system as a case study, where the system was decomposed into four functional subsystems and 12 types of components. Experimental results demonstrate that the method efficiently enables automated integration of DTs from standardized components to subsystems and the complete system. Compared with conventional monolithic modeling approaches, it significantly reduces system complexity, supports efficient component development, and accelerates system integration. For example, when the number of components exceeds 300, the proposed method generates topology connections 44.69% faster than direct information set traversal. Consequently, this approach provides a novel and effective solution to the challenges of low reusability and limited generality in DT models, laying a theoretical foundation and offering technical support for establishing a universal cross-industry DT modeling framework.

Precipitation on remote offshore islands is characterized by non-stationarity and strong stochasticity. However, rainwater harvesting systems require accurate precipitation forecasts to ensure effective water storage and discharge and a stable freshwater supply. Here, we propose an intelligent framework integrating precipitation signal decomposition, a hybrid neural network for predicting daily precipitation over the next 30 days, and digital twin technology to optimize water storage and discharge strategies for rainwater collection pools. For signal preprocessing, variational mode decomposition decomposes the original daily precipitation series into multiple intrinsic mode functions, a chaotic sparrow search optimization algorithm performs global parameter optimization, and singular spectrum analysis is applied twice to denoise high-frequency mode components. The hybrid prediction model combines a convolutional neural network to extract local temporal features, bidirectional long short-term memory to capture bidirectional long-term dependencies, and a multilayer perceptron for nonlinear mapping and regression output. Finally, this model is integrated into a digital-twin-enabled decision framework to automatically implement water storage control through water balance calculations and threshold comparisons. The proposed method achieves a test mean absolute error of 2.49 mm, a root mean square error of 5.72 mm, and an R² of 0.83; digital twin control significantly reduces overflow losses and, during typical typhoon months, maintains the water level in the rainwater collection pool below the warning threshold. In summary, we present a simulation-based pathway for optimizing small-scale intelligent rainwater harvesting systems, which offers a practical pathway to enhance the sustainability of freshwater resources on remote offshore islands by improving the efficiency of rainwater utilization and reducing dependence on external water supplies.