Multi-Strategy Coordinated Optimization of Large-Scale Crude Oil Gathering and Transportation Systems under Dynamic Conditions

To resolve the acute trade-off between energy-intensive heating for flow assurance and wax deposition risks in high-water-cut heavy oil gathering, this study proposes a synergistic optimization framework under dynamic operating conditions. The core challenge lies in solving a high-dimensional optimization problem constrained by a computationally expensive, high-fidelity “black-box” simulator and complex engineering constraints. To address this, a Phased Synergistic Memetic Algorithm (PSMA) is proposed, featuring a novel “exploration–convergence–fine-tuning” architecture. A dynamic scheduler orchestrates GA, DE, PSO, and CMA-ES operators, while an Adam-driven adaptive gradient refinement engine accelerates local convergence. Furthermore, a lexicographic fitness evaluation enforces a “safety-first” principle. Validated on two large-scale transfer stations (100 and 200 wells) in China, the simulation demonstrated high engineering credibility with mean relative errors of 2.08% and 1.67%. Importantly, the optimized scheduling eliminated prior wax deposition risks and reduced daily energy consumption by 7.42% and 8.12%, respectively. This work delivers a field-validated, robust intelligent decision tool and a generalizable framework for complex dynamic process optimization.