Optimizing Metro Station Locations and Line Layouts in Selangor using Genetic Algorithm Approach: Technical Report

Hasna Lammaihri,Marwa Erramla,Norjihan Abdul Ghani

2024-11-06

Abstract:This report presents an approach for optimizing metro station locations and line layouts in the area of Selangor, located in Malaysia. The project utilized the genetic algorithm in identifying the locations and lines layout. With population in Selangor projected to reach 7.3 million by 2024, the existing transport infrastructure is under increasing strain. This project addresses this challenge by optimizing the metro network to effectively cover the expanding population and minimize travel times. We employed a genetic algorithm to achieve these objectives, focusing on both the strategic placement of metro stations and the efficient layout of metro lines.

Computers and Society

What problem does this paper attempt to address?

The problem that this paper attempts to solve is: How to optimize the location and line layout of subway stations in Selangor, Malaysia, in order to cope with the rapidly growing population and the increasingly strained traffic infrastructure. Specifically, the research aims to improve the coverage rate and reduce passengers' travel time by optimizing the subway network.

### Problem Background

1. **Population Growth and Traffic Pressure**
- The population of Selangor is expected to reach 7.3 million by 2024.
- The existing traffic infrastructure can hardly meet the growing population needs, resulting in traffic congestion and low travel efficiency.

2. **Limitations of the Existing Traffic System**
- The current traffic system, including dense bus lines, low - frequency railway networks and crowded roads, cannot effectively meet the transportation needs in the urbanization process.

### Solution

To address the above challenges, this research uses the Genetic Algorithm (GA) to optimize the location and line layout of subway stations. Specific objectives include:

- **Maximize Coverage**: Ensure that subway stations can cover more densely populated areas and traffic - demand hotspots.
- **Minimize Travel Time**: By optimizing the line layout, passengers can reach their destinations in the shortest time.
- **Efficient Use of Resources**: While meeting transportation needs, try to reduce costs and improve operational efficiency as much as possible.

### Method Overview

1. **Data Collection and Pre - processing**
- Collect geospatial data, population density data and information on major transportation hubs (such as commercial areas, residential areas, airports, etc.).
- Use GeoJSON files to define regional boundaries and draw population density maps and major attraction center maps.

2. **Genetic Algorithm Setup**
- **Initial Population**: Randomly generate an initial population of subway station locations and line configurations.
- **Fitness Function**: Define a fitness function to evaluate the quality of each solution. The fitness function takes the following aspects into account:
- **Coverage from Population - Density Areas**:
\[
\sum\left(\int e^{-\frac{r_i^2}{\sigma^2}} \rho dA\right)
\]
- **Coverage from Demand - Generation Points**:
\[
\sum_j \sum_i e^{-\frac{r_{ij}^2}{\sigma^2}} g_j
\]
where \( r_i \) is the distance from the \( i \) - th subway station to the centroid of the area, \( \rho \) is the population density, \( \sigma \) is the standard deviation of the Gaussian function, \( A \) is the area under consideration, \( r_{ij} \) is the distance from the \( j \) - th demand - generation point to the \( i \) - th subway station, and \( g_j \) is the number of people related to the demand - generation point.

3. **Optimization Process**
- **Selection**: Select the best - performing individuals as parents for the next generation according to their fitness values.
- **Crossover**: Generate new offspring by combining the genetic information of two parents through crossover operations.
- **Mutation**: Make small - scale random changes to some individuals to introduce new solutions.
- **Elite Strategy**: Keep the best - performing individuals directly into the next generation to avoid the loss of excellent solutions.

4. **Results and Discussion**
- After 10 generations of Genetic Algorithm iterations, the optimal subway station locations and line layouts are finally determined.
- The results show that the optimized subway network significantly improves the coverage rate and reduces the average travel time.

### Conclusion

This research successfully optimizes the subway network design in Selangor through the Genetic Algorithm, providing a valuable reference for future urban traffic planning. Although there are budget limitations and computational resource constraints, this framework lays a solid foundation for follow - up research and provides directions for further improvement and application.