Quantum Key Distribution (QKD) has emerged as a robust solution for secure communications based on the principles of quantum mechanics. One of the primary challenges in QKD is minimizing the Quantum Bit Error Rate (QBER), which is affected by channel noise, hardware imperfections, and environmental disturbances. Traditional methods for QBER optimization rely on fixed threshold values and statistical models that are often inadequate under dynamic and uncertain conditions. This paper proposes an integrated approach that employs fuzzy logic neural networks (FLNNs) for real-time QBER optimization. By combining the interpretability of fuzzy logic with the learning capacity of neural networks, the proposed Adaptive Neuro-Fuzzy Inference System (ANFIS) model can predict QBER fluctuations and dynamically adjust QKD system parameters. The research proposes a mathematical formulation of the FLNN model, describes the training and validation procedures, and discusses experimental results obtained from a simulated QKD environment. The results show that the integration of neuro-fuzzy systems can significantly reduce QBER while maintaining the security integrity of QKD protocols.