Scholarlink

<p>The sustainable management of hydroelectric power consumption in intensive aquaculture operations represents a critical challenge for modern fish farming, where energy costs can constitute up to 40% of operational expenses. This study introduces a novel bio-inspired metaheuristic algorithm, the seahorse optimization algorithm with adaptive Lévy-Dunkl mutations (SOA-ALDM), specifically designed to optimize power consumption patterns in recirculating aquaculture systems (RAS). The proposed algorithm mimics the unique locomotion patterns, feeding behaviors, and territorial dynamics of seahorses (<i>Hippocampus</i> spp.), incorporating five key modifications: (1) adaptive spiral swimming mechanism for exploration–exploitation balance, (2) camouflage-based population diversity enhancement, (3) brood pouch-inspired solution storage strategy, (4) tail-grasping local intensification operator, and (5) current-riding momentum transfer mechanism. The integration of adaptive Lévy-Dunkl mutations provides superior escape capabilities from local optima through heavy-tailed statistical distributions with orthogonal polynomial foundations. The optimization framework addresses multi-objective power scheduling for aeration systems, water circulation pumps, temperature control units, and automated feeding mechanisms across 24-h operational cycles. Experimental validation was conducted on three commercial salmon farms (capacity, 150–500 tons/year) over 180-day operational periods, with performance evaluation using novel statistical metrics including the Generalized Hypervolume Indicator (GHI), Polynomial Chaos Expansion-based Sensitivity Index (PCE-SI), and Cramér-von Mises divergence measure. Comparative analysis against 12 state-of-the-art algorithms (PSO, GWO, WOA, HHO, MPA, SSA, AO, GTO, DBO, AVOA, RUN, and NGO) demonstrates that SOA-ALDM achieves superior performance with 23.7% average energy reduction, 15.3% improvement in power factor, and 89.4% Pareto front coverage and maintains optimal dissolved oxygen levels (&gt; 6.5 mg/L) while reducing peak demand charges by 31.2%. Statistical validation through Friedman ranking test (<i>p</i> &lt; 0.001) and post-hoc Wilcoxon signed-rank tests confirms significant superiority across all benchmark instances. The proposed method achieved cost savings of $47,300–$89,600 annually per facility while improving fish growth rates by 8.2% and reducing mortality by 2.3 percentage points, establishing a new paradigm for energy-efficient aquaculture operations.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Novel seahorse-inspired optimization with adaptive Lévy-Dunkl mutations for sustainable hydroelectric power management in intensive fish farming systems

  • 期刊论文

  • 发表日期：

    2026年2月16日

查看全文
  • Mohssen Elbagory,
  • Faizah Amer Altihani,
  • Hanaa Salem Marie,
  • Mostafa Elbaz

摘要

The sustainable management of hydroelectric power consumption in intensive aquaculture operations represents a critical challenge for modern fish farming, where energy costs can constitute up to 40% of operational expenses. This study introduces a novel bio-inspired metaheuristic algorithm, the seahorse optimization algorithm with adaptive Lévy-Dunkl mutations (SOA-ALDM), specifically designed to optimize power consumption patterns in recirculating aquaculture systems (RAS). The proposed algorithm mimics the unique locomotion patterns, feeding behaviors, and territorial dynamics of seahorses (Hippocampus spp.), incorporating five key modifications: (1) adaptive spiral swimming mechanism for exploration–exploitation balance, (2) camouflage-based population diversity enhancement, (3) brood pouch-inspired solution storage strategy, (4) tail-grasping local intensification operator, and (5) current-riding momentum transfer mechanism. The integration of adaptive Lévy-Dunkl mutations provides superior escape capabilities from local optima through heavy-tailed statistical distributions with orthogonal polynomial foundations. The optimization framework addresses multi-objective power scheduling for aeration systems, water circulation pumps, temperature control units, and automated feeding mechanisms across 24-h operational cycles. Experimental validation was conducted on three commercial salmon farms (capacity, 150–500 tons/year) over 180-day operational periods, with performance evaluation using novel statistical metrics including the Generalized Hypervolume Indicator (GHI), Polynomial Chaos Expansion-based Sensitivity Index (PCE-SI), and Cramér-von Mises divergence measure. Comparative analysis against 12 state-of-the-art algorithms (PSO, GWO, WOA, HHO, MPA, SSA, AO, GTO, DBO, AVOA, RUN, and NGO) demonstrates that SOA-ALDM achieves superior performance with 23.7% average energy reduction, 15.3% improvement in power factor, and 89.4% Pareto front coverage and maintains optimal dissolved oxygen levels (> 6.5 mg/L) while reducing peak demand charges by 31.2%. Statistical validation through Friedman ranking test (p < 0.001) and post-hoc Wilcoxon signed-rank tests confirms significant superiority across all benchmark instances. The proposed method achieved cost savings of $47,300–$89,600 annually per facility while improving fish growth rates by 8.2% and reducing mortality by 2.3 percentage points, establishing a new paradigm for energy-efficient aquaculture operations.