Abstract: This study addresses the multi-parameter coupled optimization problem in the bidirectional gas-water displacement development of high-angle reservoirs by constructing a field-scale well pattern model. An improved intermittent spring optimization algorithm is proposed, integrating Halton sequence initialization, tangential flight strategy, and sine-cosine disturbance mechanism. Results show that the proposed algorithm outperforms traditional methods in terms of convergence speed, global search capability, and solution stability, making it suitable for intelligent injection-production strategy design under complex constraints. Further analysis reveals that different injection methods significantly affect reservoir performance. Among them, water-alternating-gas injection achieves the best outcomes in both net present value and recovery factor; continuous gas injection shows advantages in improving recovery, while intermittent gas injection underperforms due to insufficient pressure maintenance. Additionally, the impact of objective function selection on optimization outcomes is discussed. When maximizing NPV, the optimized strategy favors early production response and rapid capital recovery; when maximizing cumulative oil production, the strategy promotes injection-production balance and displacement front advancement, contributing to enhanced ultimate recovery. This study offers new insights and effective approaches for optimizing injection-production strategies in complex displacement processes of high-angle reservoirs.