A Hybrid Framework of Neural Networks and the Sawfish Algorithm in the Context of Smart Cities for Urban Waste Management (Case Study: Ardabil)
A Hybrid Framework of Neural Networks and the Sawfish Algorithm in the Context of Smart Cities for Urban Waste Management (Case Study: Ardabil)
elham dournama1
1) B.Sc. Student of Urban Planning Engineering, Payame Noor University, Ardabil, Iran
IOI Code
:
AAHF-ABKAGE
Publication :
National Conference on Architecture and Art Technologies, Smart City, Soft and Creative Industries and Technological Culture(icact.ir)
Abstract :
Optimal urban waste management is one of the fundamental challenges in the development of smart cities, which, due to population growth, tourism fluctuations, and topographic constraints, demands innovative approaches. This study proposes a hybrid framework for intelligent waste management in the city of Ardabil. The framework is developed in two operational phases: in the first phase, a hybrid LSTM-GNN neural network model was employed to accurately predict waste generation volumes. By simultaneously accounting for temporal dependencies and spatial relationships among urban districts, this model achieved an accuracy of over 94.7%. In the second phase, the metaheuristic Sawfish Optimization Algorithm (SFO) was utilized to optimize the collection fleet routes and solve the Vehicle Routing Problem (VRP). Evaluation results indicate that the SFO algorithm exhibits higher stability and faster convergence toward optimal solutions compared to other metaheuristic methods (e.g., PSO). Analysis of various scenarios suggests that implementing this hybrid framework can improve distance traveled indices and operational costs by up to 50% compared to the current situation. The findings of this research provide a data-driven decision-support tool for urban management that, by considering the specific climatic and structural characteristics of Ardabil, enables intelligent, economical, and sustainable management of municipal services. Implementation of this framework not only enhances the quality of public services but also serves as an effective step toward reducing environmental pollution and urban traffic.
Keywords : smart waste management hybrid neural networks (lstm gnn) sawfish optimization algorithm (sfo) vehicle routing problem (vrp) smart city