Enhanced Load Forecasting and Energy Management in Grid-Connected Smart Grid Systems through the Integration of Improved Wasserstein Generative Adversarial Network and Archerfish Hunting Optimizer

Load forecasting is important for the operation and control of the power grid. However, due to its non-stationary nature and complex random characteristics, accurately predicting load behavior remains a significant challenge, such as sudden demand spikes, weather variability and changing user behavior. To overcome these difficulties, this paper proposes an innovative technique for load forecasting and energy management in grid-connected smart grid systems. The proposed approach combines the Improved Wasserstein Generative Adversarial Network (IWGAN) and the Archerfish Hunting Optimizer (AHO), which is termed the IWGAN–AHO method. The primary objective of the proposed technique is to reduce the cost of energy management and improve the forecasting performance. The IWGAN optimally predicts the load demand. The AHO is utilized to optimize the operation and scheduling of smart grid resources. The proposed method is evaluated on the MATLAB platform and demonstrates superior performance compared to existing approaches, such as Artificial Bee Colony (ABC), Genetic Algorithm (GA) and Salp Swarm Algorithm (SSA). The proposed technique attains an error of 0.04% and a computation time of 4.8[Formula: see text]s, demonstrating superior performance compared to existing techniques. These results show the higher capability of the proposed IWGAN–AHO approach in capturing the nonlinear and dynamic patterns of load demand, leading to enhanced forecasting accuracy and more efficient energy management.