بررسی و ایجاد بهترین چینش کاربری اراضی از دید حمل‌ونقل شهری با استفاده از الگوریتم فرا ابتکاری در شبیه‌ساز ترافیکی در منطقه مرکزی شهر تبریز)

ABSTRACT
The continuous increase in intra-urban travel demand and the concentration of commercial, service, and administrative activities has intensified traffic congestion in city centers, reduced road network efficiency, and exacerbated environmental impacts. Under such conditions, revising land-use allocation patterns and employing advanced optimization techniques can play a crucial role in improving urban transportation system performance. This study aims to develop an optimized model for reducing traffic congestion through land-use management in the central district of Tabriz, Iran. This area is characterized by high population density, a strong concentration of commercial and administrative functions, and a strategic position within the city’s spatial structure, making it one of the most critical traffic hotspots in Tabriz. In this research, the VISSIM microsimulation platform was used to model and evaluate urban transportation network performance under different land-use scenarios. In addition, metaheuristic algorithms, including Genetic Algorithm (GA) and Particle Swarm Optimization (PSO), were applied to identify optimal land-use allocation and spatial configuration patterns. The evaluation of the results was conducted utilizing key performance indicators such as travel time, delay, and the level of service (LOS) of road segments. The findings indicate that the proposed model significantly improves transportation network performance and reduces traffic congestion compared to the existing situation. Moreover, the employed algorithms demonstrated strong convergence capability and stability, providing efficient and reliable solutions. Overall, the results suggest that integrating spatial data, traffic microsimulation, and optimization algorithms can provide an effective framework for integrated land-use and transportation management and support the achievement of sustainable urban development goals.
Extended Abstract
Introduction
Rapid urbanization and continuous population growth in cities over recent decades have created increasingly complex challenges for urban planners, particularly in the fields of land-use allocation and transportation system management. These challenges are especially pronounced in developing countries, where unplanned urban expansion, poor infrastructure coordination, and rapid motorization have resulted in severe traffic congestion, longer travel times, environmental degradation, and declining urban livability (Norouzi, 2022). Consequently, the integration of land-use planning and transportation systems has become a critical requirement for achieving sustainable urban development (Verma et al., 2016).
Land is one of the most fundamental components of urban systems, playing a decisive role in shaping ecological balance, economic productivity, and social interactions (Koko et al., 2023). However, accelerated urban expansion and uncontrolled spatial development have significantly reduced available land resources and intensified competition among different land uses (Phan et al., 2020; Long, 2022; Besser & Hamed, 2021). According to global projections, approximately 68% of the world’s population will reside in urban areas by 2050 (United Nations Department of Economic and Social Affairs, 2019), which will further intensify pressure on urban infrastructure and transportation systems.
These demographic and spatial transformations have led to substantial changes in land use/land cover (LULC) patterns, making it essential to continuously monitor, model, and predict these dynamics employing advanced analytical tools (Xu et al., 2022; Munthali et al., 2020). In this regard, modern geospatial technologies such as Geographic Information Systems (GIS) and remote sensing have provided unprecedented capabilities for analyzing spatial patterns and guiding urban development strategies. Furthermore, emerging mobility datasets derived from digital platforms such as SafeGraph and CARTO enable researchers to capture real-world travel behavior and integrate it into urban modeling frameworks.
In parallel with data-driven approaches, optimization techniques have gained increasing attention in urban studies. These methods allow researchers to identify efficient land-use spatial configurations that minimize negative transportation outcomes while improving accessibility and network performance (Coello & Montes, 2002). Among these techniques, metaheuristic algorithms inspired by biological, physical, and evolutionary processes have proven highly effective in solving complex, nonlinear, and high-dimensional urban optimization problems (Mirjalili et al., 2014).
Metaheuristic approaches such as Genetic Algorithms (GA), Particle Swarm Optimization (PSO), Ant Colony Optimization, and other swarm intelligence methods are widely used in urban planning due to their flexibility and robustness in exploring large solution spaces (Mehrabian & Lucas, 2006). These algorithms are particularly useful for land-use optimization problems, where multiple conflicting objectives such as minimizing congestion, reducing travel distance, and improving service accessibility should be simultaneously addressed.
In this study, a hybrid optimization framework is developed to optimize land-use distribution in the central district of Tabriz, Iran. The model integrates spatial land-use data with microscopic traffic simulation outputs to evaluate the impact of different land-use configurations on urban transportation performance. The primary objective is to identify an optimal or near-optimal spatial arrangement of land uses that minimizes traffic congestion and improves overall network efficiency.
The proposed framework is designed to capture the bidirectional relationship between land-use patterns and transportation flows. While land use influences travel demand, transportation accessibility also shapes land-use development patterns. Therefore, an integrated modeling approach is necessary to accurately represent urban dynamics. The results aim to contribute to sustainable urban planning policies and provide decision-makers with a robust analytical tool for managing urban growth.
 
 
Methodology
This research is classified as applied research in terms of purpose and is based on computational modeling, simulation, and optimization techniques. The study adopts a quantitative and descriptive approach that integrates spatial data analysis, traffic simulation, and metaheuristic optimization techniques.
Study Area and Data Collection
The study focuses on Tabriz’s District 8, which represents one of the most congested and strategically significant urban areas due to its concentration of commercial, administrative, and historical land uses. The area is characterized by high population density and intense traffic demand.
Spatial datasets used in this study include land-use maps obtained from the Municipality of Tabriz and road network data extracted from OpenStreetMap (OSM). These datasets were processed and integrated into a GIS environment for further analysis and modeling.
Traffic Simulation Model
The VISSIM microscopic traffic simulation software was employed to evaluate transportation performance under different land-use scenarios. Microscopic simulation allows detailed modeling of individual vehicle behavior, including acceleration, lane-changing, car-following, and route-choice decisions.
The simulation model incorporates three levels of traffic representation: macroscopic, mesoscopic, and microscopic. Among these, microscopic simulation is selected due to its high accuracy in representing real-world traffic dynamics.
Vehicles are categorized into private and public transportation types, including taxis, buses, and emergency vehicles. Additionally, traffic demand is classified into high, medium, and low intensity groups. Peak-hour traffic conditions (morning and evening peaks) are also incorporated to ensure realistic modeling.
The performance of the transportation system is evaluated utilizing two key indicators: Vehicle Miles Traveled (VMT) and Vehicle Hours Traveled (VHT). These indicators provide quantitative measures of travel efficiency and system performance (Hillier, 1996).
Optimization Framework
In this study, two metaheuristic algorithms—Genetic Algorithm (GA) and Particle Swarm Optimization (PSO)—are employed to optimize land-use allocation. The optimization problem is formulated as a multi-objective system with the following purposes:

minimizing traffic congestion,
reducing vehicle miles traveled (VMT),
improving the level of service (LOS),
Optimizing the spatial distribution of trips,

In the GA framework, the urban land-use structure is encoded as chromosomes, where each gene represents a land-use parcel. An initial population is generated randomly, and each individual is evaluated using the traffic simulation model.
The fitness function is defined based on VMT and VHT values obtained from simulation outputs. The best-performing individuals are selected as parents, and genetic operators such as crossover and mutation are applied to generate new solutions. The evolutionary process continues until convergence criteria are satisfied.
Stopping Criteria
The algorithm terminates when stability in the objective function is observed. Specifically, after 100 iterations, the objective function values from the final 25 iterations are compared to assess stability. If no significant improvement is detected, the algorithm is stopped and the best solution is selected as the final output.
 
Results and Discussion
The simulation results indicate that the proposed optimization model significantly improves urban transportation performance in the study area. The integration of land-use data with microscopic traffic simulation enables a detailed analysis of spatial and temporal traffic patterns.
The optimized land-use configuration leads to a noticeable reduction in travel distances and traffic congestion levels. The results indicate that residential, commercial, and industrial land uses play the most influential role in shaping traffic demand patterns, with commercial land uses generating the highest level of trip attraction.
The GA-based optimization process demonstrates strong convergence behavior, improving system performance from approximately 35% in the initial state to around 85% in the optimized state. This indicates the robustness and efficiency of the proposed model in handling complex spatial optimization problems.
In early iterations, significant modifications are observed in land-use allocation, reflecting the exploratory nature of the algorithm. However, in later iterations, the system gradually stabilizes, indicating convergence toward an optimal or near-optimal solution. Residential and commercial land uses exhibit the highest frequency of relocation, confirming their strong influence on traffic generation.
A comparison with previous studies (e.g., Litman, 2011; Zegras, 2004) confirms that spatial land-use configuration has a significant impact on travel behavior and transportation efficiency. Similar to findings reported by Nasiri et al. (2019) and Scott & Rajabifard (2017), the results demonstrate that integrated land-use transportation models can substantially improve urban system performance.
Unlike many previous studies that rely on macro-level modeling approaches, this research employs microscopic traffic simulation, which provides higher accuracy and more realistic behavioral representation. Furthermore, the combined use of GA and PSO improves convergence speed and solution quality compared to single-algorithm approaches.
 
Conclusion
This study demonstrates that metaheuristic optimization techniques, particularly Genetic Algorithms and Particle Swarm Optimization, can significantly enhance urban land-use planning and transportation efficiency. The proposed model successfully reduces travel distances and improves traffic flow performance in the central district of Tabriz.
The results confirm that spatial reconfiguration of land-use patterns can lead to substantial improvements in transportation system efficiency. The optimization process achieved a performance improvement from approximately 35% to 85%, demonstrating strong convergence and model effectiveness.
However, the effectiveness of the model is influenced by the quality of input data and existing urban infrastructure conditions. In developing cities such as Tabriz, non-standardized infrastructure may limit the full realization of optimal land-use configurations. Despite these limitations, the proposed model remains robust and adaptable to different urban contexts.
Future research should focus on integrating machine learning techniques with metaheuristic optimization methods to enhance predictive accuracy and model adaptability. Additionally, incorporating socio-economic variables and long-term temporal dynamics could further improve the robustness of urban planning models.
Overall, the findings provide a comprehensive framework for integrating land-use planning and transportation modeling, contributing to sustainable urban development and improved decision-making processes.
 
Funding
There is no funding support.
 
Authors’ Contribution
The authors contributed to this manuscript in clearly defined and distinct roles. The first author was responsible for research design, data collection, data analysis, and drafting the initial version of the manuscript. The second author contributed to strengthening the theoretical framework, interpreting the results, conducting the scientific review, and performing the final editing of the manuscript.
 
 
Conflict of Interest
Authors declared no conflict of interest.
 
Acknowledgments
The authors state that no financial support was received from any individual or organization for conducting this research.