jabbrv-ltwa-all.ldf jabbrv-ltwa-en.ldf This article outlines a strategy to enhance voltage profiles in power grids by managing reactive power more efficiently through the integration of HRESin a smart grid. It highlights the challenge of voltage fluctuation in power grids owing to the erratic nature of RES like wind turbines and photovoltaic systems, which leads to an unstable voltage profile across the grid. To address these voltage fluctuations, this article proposes DSTATCOM, a reactive power compensation device to supply the necessary var to the grid. DSTATCOM helps in maintaining voltage stability, ensuring that the active power flow increases by reducing the voltage drop reliably. As a result, the overall voltage profile across the power grid improves. The core of the proposed solution involves combining Convolutional Neural Networks (CNN) with Bidirectional Long Short-Term Memory (BiLSTM) to control and optimize the performance of DSTATCOM. These advanced Artificial Intelligence (AI) techniques help manage reactive power dynamically, enhancing both the performance of DSTATCOM and the voltage profile in the grid. CNNs are used for feature extraction, while BiLSTM aids in capturing temporal dependencies in the grid’s power behavior, making this approach effective for real-time voltage regulation in smart grid environments. Also, an adaptive Parrot Optimizer (APO) is employed to fine-tune the weights of the CNN-BiLSTM network. Thus, the established system enhances voltage profile, minimizes power loss, and ensures grid stability by regulating reactive power balance. The MATLAB/Simulink environment was used to implement the proposed approach, and its performance was evaluated under three distinct scenarios.