Sign in Subscribe
Natural gas

Accurate Daily Urban Gas Load Prediction Using Genetic Algorithms

Strategic Insights

— 1 min read

Research from the PetroChina Southwest Oil and Gas Field Company, in collaboration with other institutions, has developed a novel model for predicting daily urban gas loads using genetic algorithms. This model, known as the Multiple Weather Parameter-Daily Load Prediction (MWP-DLP), was created to address the increasing demand for accurate gas load forecasting. The study's findings show that the MWP-DLP model can predict gas loads with a high degree of accuracy, with a maximum relative error of 8.2% and a mean absolute percentage error (MAPE) of 2.68%. This breakthrough has significant implications for gas companies, as it enables them to make informed decisions about peak shaving schemes and natural gas reserves.

Key Takeaways:

  • Researchers at PetroChina Southwest Oil and Gas Field Company and other institutions have developed a novel model for predicting daily urban gas loads using genetic algorithms.
  • The Multiple Weather Parameter-Daily Load Prediction (MWP-DLP) model considers factors such as average temperature, solar radiation, cumulative temperature, wind power, and temperature change of the building foundation.
  • The genetic algorithm used in the MWP-DLP model solves the forecasting equation, resulting in accurate predictions of gas loads.
  • The study's results showed a consistent trend between the predicted and actual gas loads, with a maximum relative error of 8.2% and a mean absolute percentage error (MAPE) of 2.68%.
  • The MWP-DLP model has practical significance for gas companies, enabling them to make informed decisions about peak shaving schemes and natural gas reserves.
  • The feasibility of the MWP-DLP prediction model was verified, demonstrating its effectiveness in predicting daily urban gas loads.

Statistics:

  • Maximum relative error: 8.2%
  • Mean absolute percentage error (MAPE): 2.68%
  • Daily gas load predictions using the MWP-DLP model accurate to within 8.2% of actual load

Sources:

  • A Novel Model for Accurate Daily Urban Gas Load Prediction Using Genetic Algorithms. Algorithms, 2025,18(6):347.
  • MDPI AG (publisher of Algorithms)
  • NewsRx (2025). Data from PetroChina Southwest Oil and Gas Field Company Update Knowledge in Algorithms (A Novel Model for Accurate Daily Urban Gas Load Prediction Using Genetic Algorithms). Life Science Weekly. July 8, 2025; p 580.

Read more