Forecast of energy use made easy in Andhra with the help of artificial intelligence

Andhra Pradesh, according to APTRANSCO officials, is the first State in India to have developed ‘the most accurate’ day-ahead forecast model using ML and AI. 

Published: 28th December 2020 07:46 AM  |   Last Updated: 28th December 2020 07:46 AM   |  A+A-


(Representational Image)

By Express News Service

VIJAYAWADA: The State Load Dispatch Centre (SLDC) of the Transmission Corporation of Andhra Pradesh Ltd (APTRANSCO) has developed a day-ahead electricity forecasting model using machine learning (ML) and artificial intelligence (AI) to meticulously plan power management, which helps minimise power purchase costs. Andhra Pradesh, according to APTRANSCO officials, is the first State in India to have developed ‘the most accurate’ day-ahead forecast model using ML and AI. 

With the model yielding accurate results in forecasting next day’s electricity consumption, including day-ahead electricity demand for every 15 minutes, the officials said forecasting models for wind/solar generation, grid frequency, market price, surplus NTPC/Central Generating Stations power at national level will also be developed. 

“Google with World Bank’s support has offered to jointly develop the forecasting model with the AP State Load Dispatch Centre and the SLDC has done the entire development in-house. This will be one of the major steps towards reaching the objective of State government to make AP destination of cost-effective power,” the officials said, in a statement on Sunday.

They said forecasting is essential to plan how much energy should be dispatched from each generation unit - thermal, gas or renewable - and estimate market purchases in advance. Manual forecasting is not accurate and leads to over drawl/under drawl of power. 

This will make the grid management a tedious task as unforeseen situations like outages, sudden fall of wind/solar power and others will have to be met from real time market, which will be expensive. To overcome these challenges, the SLDC developed a model by using 25 years of data, including climate data, holidays, Covid-19 lockdown and seasonal information, weather forecast etc. 

The data flow into its servers on a real-time basis, helps achieve a high degree of accuracy with an error of less than 3 per cent in forecasting.“Electricity demand forecast is a big task before the SLDC to prepare for purchase or sale of power the next day. The load-demand mismatch leads to over or under drawl from the national grid, which entails heavy penalties and sometimes lead to power cuts as well,” the SLDC officials said.

In fact, electricity price forecast has become increasingly important worldwide with the transition of the electric utility industry from a long term conventional PPA era to a short term renewable dominant market economy. Power generators and Discoms rely on price forecasting information to develop their corresponding bidding strategies. If a generator has an accurate forecast of the prices, it can develop a bidding strategy to maximise its profit. On the other hand, a Discom can make a plan to minimise its own electricity cost if an accurate price forecast is available for the next day, the officials added. 
Now, the SLDC is in the process of developing a model for dispatch scheduling.


Disclaimer : We respect your thoughts and views! But we need to be judicious while moderating your comments. All the comments will be moderated by the editorial. Abstain from posting comments that are obscene, defamatory or inflammatory, and do not indulge in personal attacks. Try to avoid outside hyperlinks inside the comment. Help us delete comments that do not follow these guidelines.

The views expressed in comments published on are those of the comment writers alone. They do not represent the views or opinions of or its staff, nor do they represent the views or opinions of The New Indian Express Group, or any entity of, or affiliated with, The New Indian Express Group. reserves the right to take any or all comments down at any time.

flipboard facebook twitter whatsapp