Page 214 - 2024-Vol20-Issue2

210 | Gochhait, Sharma & Bachute

TABLE I.
COMPARATIVE LOAD ANALYSIS FOR ODISHA AND DELHI

Region Period Interval Seasons Load Data
Odisha January 2019 - December 2022 1-hour intervals Spring Load demand in MW
Summer Load demand in MW
Delhi January 2019 - December 2022 24-hour intervals Load demand in MW
Fall Load demand in MW
Winter Load demand in MW
Spring Load demand in MW
Summer Load demand in MW
Load demand in MW
Fall
Winter

Fig. 1. Proposed approaches and load forecasting models.

(OPTCL) in effectively managing the electricity load during throughout the year, considering the seasonal variations. The
different seasons. The study results can assist OPTCL in ac- study findings can be utilized to accurately predict the elec-
curately predicting the electricity demand in Odisha during tricity demand during each season and implement appropriate
each season and implementing necessary measures to ensure measures to ensure a consistent and uninterrupted power sup-
an uninterrupted power supply to customers. By leveraging ply to customers. By doing so, the overall reliability and
this information, OPTCL can enhance the overall reliability efficiency of the power grid in Delhi can be improved.
and efficiency of the power grid in Odisha.
Similarly, the research paper also investigates the electricity D. Proposed Approaches and Load Forecasting Models
load data for Delhion a monthly basis, from January 2019 to This research paper presents a comprehensive six-step method-
December 2022, with measurements taken at 24-hour inter- ology for the development of an efficient load forecasting
vals. The objective is to calculate the electricity load demand strategy using the 1D CNN BI LSTM methodology. The out-
in MW and analyze the seasonal variations in consumption. lined steps are as follows:
The results reveal distinctive load demand patterns across the Step 1: Data Collection: The initial phase involves the gath-
four seasons: Spring (March to May), Summer (June to Au- ering of two distinct datasets, comprising historical load de-
gust), Fall (September to November), and Winter (December mands and past weather data. These datasets serve as the
to February). foundation for a comprehensive analysis.
The insights gained from this study have significant implica- Step 2: Model Selection: Careful consideration is given to
tions for effectively managing the electricity load in Delhi selecting appropriate machine learning models for load fore-