Scholarlink

<p>The increasing worldwide need for sustainable agriculture has prompted using state-of-the-art technology to improve conventional agricultural methods. This study investigates the deployment of Internet of Things (IoT)-enabled smart irrigation systems enhanced by intelligent machine learning algorithms to enhance water efficiency in agriculture. With a network of sensors monitoring soil moisture, weather, and temperature, in the present paper we propose Bi-Long Short-Term Memory (BiLSTM) model along with a modified optimum Golden Search Optimization (GSO) to make real-time irrigation decisions. The proposed GSO-BiLSTM model can predict and advise crops using the IoT-based smart irrigation recommendation issues. The BiLSTM crop prediction approach utilizes the GSO algorithm for preprocessing and feature selection. The climate data utilized in this study includes many parameters that impact rainfall and agricultural productivity. The input is then improved via pre-processing. The GSO approach employs a selection process to identify the most relevant features, resulting in enhanced prediction accuracy and expedited implementation. The dataset's most relevant characteristics are selected based on the optimal fitness values using BiLSTM for crop projection. The accuracy of GSO-BiLSTM is 10.79%, 12.14%, 5.44%, 3.82%, and 2.89% better than SVR + K-mean, LSTM-GBT, IDCSO-WLSTM, DliSA, and IoTDL-SIS.</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

GSO-BiLSTM: performance assessment of IoT enable smart irrigation system using intelligent algorithms

  • 期刊论文

  • 发表日期：

    2025年8月4日

查看全文
  • Sandeep Bhatia,
  • Zainul Abdin Jaffery,
  • Shabana Mehfuz

摘要

The increasing worldwide need for sustainable agriculture has prompted using state-of-the-art technology to improve conventional agricultural methods. This study investigates the deployment of Internet of Things (IoT)-enabled smart irrigation systems enhanced by intelligent machine learning algorithms to enhance water efficiency in agriculture. With a network of sensors monitoring soil moisture, weather, and temperature, in the present paper we propose Bi-Long Short-Term Memory (BiLSTM) model along with a modified optimum Golden Search Optimization (GSO) to make real-time irrigation decisions. The proposed GSO-BiLSTM model can predict and advise crops using the IoT-based smart irrigation recommendation issues. The BiLSTM crop prediction approach utilizes the GSO algorithm for preprocessing and feature selection. The climate data utilized in this study includes many parameters that impact rainfall and agricultural productivity. The input is then improved via pre-processing. The GSO approach employs a selection process to identify the most relevant features, resulting in enhanced prediction accuracy and expedited implementation. The dataset's most relevant characteristics are selected based on the optimal fitness values using BiLSTM for crop projection. The accuracy of GSO-BiLSTM is 10.79%, 12.14%, 5.44%, 3.82%, and 2.89% better than SVR + K-mean, LSTM-GBT, IDCSO-WLSTM, DliSA, and IoTDL-SIS.