Scholarlink

<p>Maintaining healthy aquatic ecosystems, guaranteeing sustainable farming methods, avoiding fish mortality, and fulfilling safety regulations for aquaculture and public health all depend on the ability to predict ammonia toxicity in aquaculture water. The dense systems of inland ponds used for aquaculture in Andhra Pradesh’s western delta have resulted in a highly contaminated organic effluent load, primarily in terms of BOD, ammonia, nitrates, and other pollutants like calcium, potassium, sodium, and chlorides. The water condition of aquaculture waterways in various sites is examined in the current investigation. A significant portion of the samples were of extremely low quality, with ammonia contents higher than the safety thresholds set by health regulations. The research study offers a sophisticated intelligent soft computing method for ammonia level prediction that takes into account the potential toxicity of ammonia.&#xa0;This method integrates iterative stepwise robust model-based imputation for preprocessing, the Hybrid Prairie Dog Optimization Algorithm with Binary Light Spectrum Optimizer (PDOA-BLSO) has been used for feature selection, and Stereoscopic Scalable Quantum Convolutional Neural Networks with Simplicial Attention Networks (SSQC-SAN) are utilized for prediction. The Emperor Penguin Optimizer (EPO) was utilized as well to enhance optimization. The hybrid approach proposed here indeed demonstrated great accuracy in predicting, with performance metrics above 99%, providing a reliable, easy-to-implement tool for stakeholders and policymakers to monitor ammonia contamination in real time to support sustainable aquaculture management.</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

Enhanced ammonia contamination prediction in aquaculture ponds using stereoscopic scalable quantum convolutional networks with simplicial attention network

  • 期刊论文

  • 发表日期：

    2025年10月14日

查看全文
  • Mahadev Kalyanrao Patil,
  • Pritam Pravin Kothari,
  • Srilakshmi ramya Sakamudi,
  • K. Kalaivanan

摘要

Maintaining healthy aquatic ecosystems, guaranteeing sustainable farming methods, avoiding fish mortality, and fulfilling safety regulations for aquaculture and public health all depend on the ability to predict ammonia toxicity in aquaculture water. The dense systems of inland ponds used for aquaculture in Andhra Pradesh’s western delta have resulted in a highly contaminated organic effluent load, primarily in terms of BOD, ammonia, nitrates, and other pollutants like calcium, potassium, sodium, and chlorides. The water condition of aquaculture waterways in various sites is examined in the current investigation. A significant portion of the samples were of extremely low quality, with ammonia contents higher than the safety thresholds set by health regulations. The research study offers a sophisticated intelligent soft computing method for ammonia level prediction that takes into account the potential toxicity of ammonia. This method integrates iterative stepwise robust model-based imputation for preprocessing, the Hybrid Prairie Dog Optimization Algorithm with Binary Light Spectrum Optimizer (PDOA-BLSO) has been used for feature selection, and Stereoscopic Scalable Quantum Convolutional Neural Networks with Simplicial Attention Networks (SSQC-SAN) are utilized for prediction. The Emperor Penguin Optimizer (EPO) was utilized as well to enhance optimization. The hybrid approach proposed here indeed demonstrated great accuracy in predicting, with performance metrics above 99%, providing a reliable, easy-to-implement tool for stakeholders and policymakers to monitor ammonia contamination in real time to support sustainable aquaculture management.