Scholarlink

<p>The quality of electricity consumption forecasting is crucial for the planning of Pakistan's power system. In order to improve the reliability of prediction results, this paper proposes a combined prediction model based on long short-term memory and Monte Carlo simulation. Firstly, the factors influencing Pakistan's electricity consumption were analyzed, such as economic growth rate, population size, and industrial development. Second, a deep-learning technique based on a long short-term memory (LSTM) neural network machine-learning model to forecast electricity consumption in Pakistan. Then, the probabilistic analysis of Monte Carlo with a 95% confidence interval forecasts the future consumption of electricity. At last, the historical data of Pakistan’s GDP, Population, industry efficient and electricity consumption are used to testified the efficient, operative, consistent, and robust of the proposed model. The research of this paper is helpful to policymakers in Pakistan as it significantly contributes to more effective electricity management policies, energy supply industry behavioral changes, and reduced energy consumption.</p>

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

Forecasting of electricity consumption in Pakistan based on integrating machine learning algorithms and Monte Carlo simulation

  • 期刊论文

  • 发表日期：

    2025年1月2日

查看全文
  • Muhammad Umair Nazir,
  • Jinchao Li

摘要

The quality of electricity consumption forecasting is crucial for the planning of Pakistan's power system. In order to improve the reliability of prediction results, this paper proposes a combined prediction model based on long short-term memory and Monte Carlo simulation. Firstly, the factors influencing Pakistan's electricity consumption were analyzed, such as economic growth rate, population size, and industrial development. Second, a deep-learning technique based on a long short-term memory (LSTM) neural network machine-learning model to forecast electricity consumption in Pakistan. Then, the probabilistic analysis of Monte Carlo with a 95% confidence interval forecasts the future consumption of electricity. At last, the historical data of Pakistan’s GDP, Population, industry efficient and electricity consumption are used to testified the efficient, operative, consistent, and robust of the proposed model. The research of this paper is helpful to policymakers in Pakistan as it significantly contributes to more effective electricity management policies, energy supply industry behavioral changes, and reduced energy consumption.