Application of Hybrid Intelligent Systems for Forecasting and Managing Sustainable Development of the Urban Environment Based on Fuzzy Neural Networks

This study presents a hybrid intelligent forecasting and decision-support framework for sustainable urban development based on the Adaptive Neuro-Fuzzy Inference System (ANFIS). The research employs data from 12 major Russian cities, covering environmental, social, and economic indicators across a nine-month period in 2025. The proposed architecture integrates fuzzy logic with neural network learning, enabling adaptive modeling of complex urban dynamics. A hybrid optimization algorithm combining genetic search and backpropagation was used to calibrate model parameters and enhance predictive accuracy. Evaluation through MAPE, RMSE, R², and Theil’s U statistic demonstrated superior performance of the ANFIS model compared to conventional approaches such as Random Forest, ARIMA, and linear regression. The results confirm the effectiveness of hybrid intelligent systems in supporting strategic planning and real-time management in smart city environments, providing a scalable foundation for sustainable urban governance.