Multiobjective Optimal Design and Application of Large-Scale Self-Pressurized Drip Irrigation Network Systems

Owing to the challenges posed by large-scale drip irrigation pipe networks (i.e., multiple hierarchical levels, complex layout configurations, and extended pipeline lengths), this study proposes a network partitioning method based on three fundamental units: level, branch, and segment. Coupled with a dynamic pipe diameter adjustment mechanism, this approach allows for precise modifications to individual pipeline sections, effectively overcoming the design limitations associated with traditional preset diameter-changing nodes. Building on this foundation, a universal multiobjective optimization mathematical model is developed for gravity-fed drip irrigation systems, comprehensively accounting for both economic efficiency (minimizing network investment) and reliability (minimizing average node surplus water head). To solve this optimization problem, the study integrates the energy valley optimizer (EV) algorithm with the Harmony search (HS) algorithm, proposing a new hybrid energy valley–Harmony search (HEV-HS) algorithm. By combining global exploration with local refinement, the proposed method significantly improves multiobjective optimization efficiency. Simulation results demonstrate that, compared with traditional algorithms such as particle swarm optimization and genetic algorithm, HEV-HS achieves an 18.2%–25.6% faster convergence rate, confirming its superiority in multiobjective optimization. When applied to a large-scale gravity-fed drip irrigation project in Xinjiang, the optimized solution reduces the total network investment from the original design’s RMB 122.95 million to RMB 113.66 million (a 7.56% reduction), and decreases the average node surplus water head from 11.23 to 9.33 m (a 16.92% reduction), validating the effectiveness and applicability of the proposed optimization model.