Short-term global horizontal irradiance forecasting using weather classified categorical boosting

Ahmed, U., Khan, A. R., Mahmood, A., Rafiq, I., Ghannam, R. and Zoha, A. (2024) Short-term global horizontal irradiance forecasting using weather classified categorical boosting. Applied Soft Computing, 155, 111441. (doi: 10.1016/j.asoc.2024.111441)

Abstract

Accurate short-term solar irradiance (SI) forecasting is crucial for renewable energy integration to ensure unit commitment and economic load dispatch. However, hourly SI prediction is very challenging due to atmospheric conditions and weather fluctuations. This study proposes a hybrid approach using weather classification and boosting algorithms for short-term global horizontal irradiance (GHI) forecasting. In data pre-processing steps, we employ random forest for feature selection and K-means clustering for weather classification. The weather-based clustered data is used for the model training using categorical boosting (CatBoost). The proposed weather-classified categorical boosting (WC-CB) scheme is compared with benchmarks in literature like adaptive boosting (AdaBoost), bi-directional long short-term memory (BiLSTM) and gated recurrent unit (GRU) using datasets from two distinct geographical locations obtained from the National Solar Radiation Database (NSRDB). The results show that the proposed WC-CB hybrid approach has lower forecast errors compared to conventional CatBoost modelling. The error reduction of 16% and 39% in root mean square error and 6% and 9% in mean absolute error is recorded for the two datasets, respectively. These findings demonstrate the importance of weather classification in improving forecasting accuracy with potential implications for broader renewable energy applications.

• Item Type: Articles
• Status: Published
• Refereed: Yes
• Glasgow Author(s) Enlighten ID: Zoha, Dr Ahmed and Ghannam, Professor Rami and Khan, Ahsan Raza

Creator Roles

Khan, A. R.: Writing – review and editing, Writing – original draft, Methodology, Formal analysis, Conceptualization
Ghannam, R.: Writing – review and editing, Conceptualization
Zoha, A.: Writing – review and editing, Supervision, Project administration, Conceptualization

• Authors: Ahmed, U., Khan, A. R., Mahmood, A., Rafiq, I., Ghannam, R., and Zoha, A.
• College/School: College of Science and Engineering > School of Engineering > Autonomous Systems and Connectivity; College of Science and Engineering > School of Engineering > Electronics and Nanoscale Engineering
• Journal Name: Applied Soft Computing
• Publisher: Elsevier
• ISSN: 1568-4946
• ISSN (Online): 1872-9681
• Published Online: 11 March 2024
• Copyright Holders: Copyright © 2024 The Author(s)
• First Published: First published in Applied Soft Computing 155:111441
• Publisher Policy: Reproduced under a Creative Commons license