To address the issue that existing non-intrusive Embedded Discrete Fracture Model (EDFM) methods tend to cause complex grid topology disorder and damage the native data structure of reservoir simulators due to the introduction of non-neighboring connections (NNCs), this paper proposes a non-intrusive EDFM simulation approach that directly modifies the dual-medium model without adding NNCs. The proposed method can construct an equivalent dual-medium model with the same grid pore volume, inter-grid transmissibility, and well index as the EDFM model, thereby enabling equivalent EDFM computations directly within commercial reservoir simulators. Taking immiscible two-phase flow as an example, the method is verified through multiple cases with different fracture distributions (linear fractures, intersecting fractures, complex fracture networks) and development modes (depletion development, waterflooding development). Numerical results demonstrate that the proposed method is highly consistent with the standard EDFM in terms of oil production rate, water cut, pressure, and oil saturation distribution, with an average computational error below 0.3%. Therefore, the proposed approach combines the high-precision characterization capability of EDFM for complex fracture networks and the engineering practicability of the dual-medium model, providing a new route for efficient numerical simulation of fractured oil reservoirs.

Precise forecasting of the initial productivity rates of infill wells is essential for the effective exploitation of offshore reservoirs characterized by high water-cut. However, conventional reservoir simulation and basic machine learning models often suffer from high computational complexity and low interpretability. This research introduces a hybrid data-driven framework that combines ensemble feature engineering with a random forest model optimized through the crayfish optimization algorithm. The primary controlling factors were identified through a majority voting mechanism involving five feature selection algorithms. Subsequently, the COA was utilized to optimize the parameters of the random forest algorithm to improve its predictive robustness. The proposed EFE-COA-RF model achieves a testing MAE of 6.831 and an R² of 0.954, outperforming standard machine learning models and other optimization-based variants. The complete training process requires approximately 10.8 min, whereas the prediction time for the testing set is approximately 0.03 s. These results demonstrate that the proposed framework provides an accurate, interpretable, and efficient tool for rapid productivity evaluation in mature offshore oilfields.

Tuberculosis (TB) remains a major global public health priority and continues to prevail in many regions. This study takes Jiangsu Province as the research area to analyze the impacts of population mobility and seasonal variations on TB transmission, identify key epidemic risk factors, and select the most cost-effective prevention and control interventions. According to the epidemiological status and characteristics of TB in Jiangsu Province, four seasonal multi-regional $SCLIT$ compartmental models were established. The models incorporate seasonal effects, intra- and inter-provincial population mobility, fast and slow disease progression, reinfection, incomplete treatment and vaccination. We calibrated model parameters using notified TB case data and population mobility records from 2009 to 2019, and applied Akaike and Bayesian information criteria ($AIC$ and $BIC$) to select the best-fitting model. Sensitivity analysis was performed for the average basic reproduction number \(\bar{\mathcal{R}}_0\) and the size of infected populations. In addition, cost-effectiveness analysis was conducted based on the actual expenditure data of local TB prevention and control. Periodic transmission exerts a more prominent impact on pulmonary tuberculosis prevalence than population mobility. The average basic reproduction number of TB in Jiangsu Province is estimated to be 3.5633 (95\% CI: 3.4988–3.6278), demonstrating that the local TB epidemic presents prominent annual periodic fluctuations with a cycle of approximately 12.2068 months. Model predictions of overall TB trends in Jiangsu from 2023 to 2025 are generally consistent with real-world observations. Compared with the baseline scenario, the optimal intervention strategy can avert 16,380 infections (1.7\%) and 62,880 deaths, and reduce 3,269,500 disability-adjusted life years (DALYs). Meanwhile, total costs are cut by 461.65 million RMB (14.4\%). Baseline projections based on the 2015 incidence rate of 37.99 per 100,000 population show that the provincial TB incidence would decrease by only 26.0\% by 2025 and 28.5\% by 2035. These values are considerably lower than the WHO End TB Strategy targets, which require a 50\% incidence reduction by 2025 and a 90\% incidence reduction by 2035 relative to the 2015 baseline. A reasonable combination of multiple interventions can effectively reduce the TB disease burden. The findings provide quantitative references for Jiangsu and other regions with moderate TB burden to strengthen TB control and accelerate the achievement of WHO strategic goals.