Analysis of recharging station location optimization for e-scooters through micro-simulation

Abstract

The concept of micro-mobility has become increasingly common in our lives with the introduction of electric scooters due to the advancement of technology and rapid urbanization. The traffic congestion caused by the increasing population and density, especially in metropolitan cities, has compelled people who seek mobility throughout the day to turn to alternative transportation modes. Individuals who need to reach the main backbone of the city, such as railways or piers, often prefer public transportation modes like buses or minibuses, or they choose to walk. However, it is a well-known fact that passengers are not particularly satisfied with these modes of transportation due to factors such as fixed departure schedules or traffic congestion along the routes. Especially during and after the global pandemic, which had a worldwide impact, there has been a decrease in the demand for public transportation modes known to carry risks in terms of hygiene and health conditions, and passengers have turned to passenger cars for travel. This situation contributes to the environmental effects already influenced by the transportation sector, such as air pollution, changes in surface temperatures that lead to climate change, and excessive consumption of limited energy resources. With advancing technology, alternative fuel vehicles are becoming increasingly prevalent, and there is a global increase in the number of users. The use of renewable energy in these vehicles helps reduce the negative environmental impacts caused by conventional motor vehicles, and it is believed that in the future, they will serve the transportation sector to a level that eliminates these effects entirely. Electric scooters, which consume alternative fuel like electric vehicles, are particularly appealing to passengers due to their easy accessibility compared to other modes of transportation, their suitability for short-distance travel, and their small size. They are commonly used in areas with traffic congestion and for travel between transportation modes or home-to-station journeys. However, the biggest challenge for electric vehicles and scooters is the need for battery charging. The distance that can be covered with a fully charged battery varies depending on factors such as battery specifications, road conditions, vehicle features, and more. In this context, users choose to travel with electric scooters instead of electric vehicles to be less affected by traffic congestion. They can also use electric scooters to travel from origin to destination by combining them with public transportation vehicles. Passengers who do not own personal electric scooters can rent vehicles provided by electric scooter operators to perform their journeys. Operators either use a vehicle to travel within their service area to collect and distribute scooters for battery charging or simply make brief stops for battery replacement. It should be noted that the energy consumed by the vehicles used during these operations has a negative impact on the environment and contributes to traffic congestion. Within the constraints of transportation vehicles, vehicle routing problems have emerged to find the most optimal routes between predetermined points in a known cluster of cities. With the development of technology and evolving needs, new constraints have been added, giving rise to sub problems such as vehicle routing problems with time windows, pickup and delivery vehicle routing problem, multiple depot vehicle routing problem, facility location problem, dial-a-ride vehicle routing problem, and electric vehicle routing problems. Depending on a specific problem, constraints such as time limitations, picking up and dropping off items from/to different nodes, determining depot locations, and customer collection are considered, and various solution methods are tested based on the size of the problem. In the context of electric vehicle routing problems, considerations include achieving the minimum energy consumption, the minimum cost, and the shortest distance. Constraints such as the driving range, location of recharging stations, time spent at recharging stations, battery specifications, and geographical characteristics of the sample area are taken into account. This study focuses on the widespread use of electric scooters and aims to realistically calculate the energy consumption of electric scooters with given model which also helps to determine the optimal locations for charging stations among potential points on travel routes, considering the needs of both electric scooter operators and individual users. By considering nodes that are treated as origin, destination, or possible recharging station points, and utilizing the road network that connects these points, the Recharging Station Locating Problem with Energy Consumption is formulated. This problem incorporates three-dimensional road geometry and realistic electric scooter energy consumption approach to determine the optimal locations for recharging stations with the aim of minimum cost. Solutions to the problem formulated are obtained by IBM ILOG CPLEX 12.9.0 in order to determine the location of recharging stations. To evaluate the proposed problem, Istanbul Technical University, Ayazaga Campus, is chosen as the study area. The study area consists of 30 travel routes, 30 nodes, and 76 edges. The maximum slope in the area is calculated to be 8.5%. In the scope of the study, realistic energy consumption approach is considered along with distance-based fixed energy consumption approaches. The need for recharging stations on the nodes for different energy consumption approaches are compared and the outcome showed the possibility of running out of energy before ending the journey. Analyses have indicated the impact of energy consumption calculations on the number of charging stations and consequently on the total cost, as well as the possibility of some electric scooters running out of energy. The network performances are evaluated for different scenarios through the Eclipse SUMO simulation software, which utilizes electric scooter travel data and vehicle demand data obtained from field observations. In this context, performance evaluations using optimization outputs are carried out by comparing travel times for different energy consumption approaches through micro-simulation with only electric scooters. The same approach is applied to mixed traffic scenarios with electric scooters, where the impacts of different energy consumption approaches on network performance, including travel times, are evaluated. The samples are also used to measure the effects of different energy consumption approaches in mixed traffic on electric scooter travel and to assess the impact of traffic congestion on electric scooter journeys. The present study focuses on the routing of electric scooters, optimization of recharging station locations using realistic and distance-based energy consumption approaches, and evaluating the solutions obtained by considering electric scooters both in mixed traffic and separately through various scenarios through micro-simulation program. The impact of different energy consumption approaches of electric scooters on network performances are discussed. The results highlight the importance of accurately determining the energy consumption of electric scooters for the proper identification of charging restation locations. Future extensions to the research involves considering larger networks to validate the proposed model, and improving the efficiency of model perfomance by employing heuristics solution methods.