Scholarlink

In a variety of fields including climatology, energy, and engineering, precise solar irradiance measurement is crucial. For most studies, the model outputs are the irradiance values, and the model inputs are meteorological parameters. Although there have been many recent developments in solar irradiation techniques, computation error and achieving high prediction accuracy continue to be major challenges. In this work, a novel Adam Lyrebird Optimization Algorithm_ Deep Long Short Term Memory (ALOA_DLSTM) method for solar irradiance prediction is proposed. Primarily, input time series solar irradiance data is gained from the database. Further, technical indicator extraction is carried out, where indicators, like Relative Strength Index (RSI), Linear Regression Forecast (LRF), Simple Moving Average (SMA), and Weighted Moving Average (WMA) are extracted. Then the solar irradiance prediction is carried out employing DLSTM trained with the proposed ALOA. Moreover, ALOA is introduced by integrating Adam Optimizer and the Lyrebird Optimization Algorithm (LOA). Furthermore, the supremacy of proposed ALOA_ DLSTM is investigated concerning Root Mean Square Error (RMSE), Mean Average Percentage Error (MAPE), MSE, and Relative Absolute Error (RAE) and is found to have gained values of 0.270, 0.138, 0.073, and 0.203.

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

Adam Lyrebird Optimization-Based DLSTM for Solar Irradiance Prediction Using Time Series Data

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Vikas Gautam,
  • Banalaxmi Brahma

摘要

In a variety of fields including climatology, energy, and engineering, precise solar irradiance measurement is crucial. For most studies, the model outputs are the irradiance values, and the model inputs are meteorological parameters. Although there have been many recent developments in solar irradiation techniques, computation error and achieving high prediction accuracy continue to be major challenges. In this work, a novel Adam Lyrebird Optimization Algorithm_ Deep Long Short Term Memory (ALOA_DLSTM) method for solar irradiance prediction is proposed. Primarily, input time series solar irradiance data is gained from the database. Further, technical indicator extraction is carried out, where indicators, like Relative Strength Index (RSI), Linear Regression Forecast (LRF), Simple Moving Average (SMA), and Weighted Moving Average (WMA) are extracted. Then the solar irradiance prediction is carried out employing DLSTM trained with the proposed ALOA. Moreover, ALOA is introduced by integrating Adam Optimizer and the Lyrebird Optimization Algorithm (LOA). Furthermore, the supremacy of proposed ALOA_ DLSTM is investigated concerning Root Mean Square Error (RMSE), Mean Average Percentage Error (MAPE), MSE, and Relative Absolute Error (RAE) and is found to have gained values of 0.270, 0.138, 0.073, and 0.203.