Multi Objective Fuzzy Network Based Big Data Analytics System for Energy Consumption Prediction in Smart Grid

In the modern era, smart grids are generates vast amount of heterogeneous data, which are making the challenges for effective energy management systems. Consequently, the prediction of energy consumption is important for reducing the operational cost, ensuring grid stability and supporting maintainable energy usage. Moreover, the study of energy usage in buildings is popular due to its significant environmental and financial implications. Smart metering can improve building energy efficiency and management. However, the existing prediction strategies are failed to manage the large scale energy data, dynamic behaviour of grid data and uncertainty due to the diversity of family home types and the non-linear trends. Therefore, a specialized models that consider both temporal and spatial factors are required. This study aims to address these issues through a Multi-objective Golden Eagle Optimization-based (MGEO) hybrid deep dynamic convolutional fuzzy network (HDDCFN) method. Multi-objective optimization is used to optimize the hyperparameters of the algorithm and prediction model. Regarding historical energy usage and weather conditions, the suggested method incorporates a combination of linear and non-linear predictions. Here, fuzzy logic is used to manage the inaccurate and uncertainty data while applying the Multi-objective Golden Eagle Optimization model. Here, large range of data were processed with big data analytics and provides adaptive and reliable prediction outcomes. Moreover, the developed approach improves the energy distribution, load forecasting and demand response performance. The simulation was done using MATLAB and the developed system has demonstrates lower Mean Absolute Error (MAE) of 0.24, decreases the Mean Square Error (MSE) of 0.27 and enhance the R2 of 0.92 while comparing the traditional models.