Advancing Energy Management: A Comprehensive Analysis of Electricity Consumption Forecasting and Trends in Pakistan using ANFIS

Electricity consumption (EC) forecasting is crucial for effective planning and operation in power system energy management, enhancing energy efficiency and system reliability. The challenges in power systems arise from their vast complexity, widespread distribution, and susceptibility to unexpected events. Current research addresses these complexities by considering demographic factors such as weather, climate, and load demands. Utilizing data from the different literature sources spanning a twenty years period (1991-2020) on a monthly basis, Adaptive Neuro-Fuzzy Inference System (ANFIS) module is developed for long-term load forecasting. The simulations were conducted using the MATLAB software and designing ANFIS module. The study also investigated the impact of weather variables, including temperature, humidity, wind speed, rainfall, actual load, and previous load on load forecasting. Unlike traditional statistical or machine learning models, the ANFIS approach effectively captures nonlinear dependencies and regional disparities in electricity consumption. Implemented using MATLAB, the model demonstrates prediction accuracy above 91% across provinces. Performance evaluations revealed that ANFIS predict the electricity consumption with high accuracy. Current research will increase the reliability and performance of computer-based performance models at the estimation of power generation stage. Also, present simulation based predictive analysis will provide understanding to the researchers and policy makers in the power sector about the user function of ANFIS.