GMM-P: Intelligent Method Based a Mobile Battery Swapping Service for Electric Vehicles

Abstract

The effect of greenhouse gas (GHG) from fossil fuels is a critical global issue. Thus, the trend of electric vehicles (EVs) is rapidly increasing to reduce the GHG effect. The development of EV facilities such as battery charging stations (BSCs) and fast charging techniques is the main priority to deal with the demand of EV users. One of the solutions is the battery swapping station (BSS) because this station replaced the empty battery with the fully charged battery. Therefore, their EV users do not need to wait for battery charging. However, the number of BSSs is not enough to facilitate and cover their users. Then, the researcher proposed the concept of a mobile BSS in terms of van and taxi. The difference between the battery swapping van (BSV) and battery the swapping taxi is the number of carried batteries. The BSV can carry more batteries than a taxi. The concept of our proposed battery swapping service is that the BSV departs from the BSS with fully charged batteries to EV users who request the battery replacement. After serving all EV requests, BSV returns to BSS to charge the collected exhausted batteries. The challenge of this service is the searching process for the optimal and shortest delivery path. Therefore, this paper proposed an intelligent method using the Gaussian mixture model (GMM) and particle swarm optimization (PSO) named 'GMM-P'. The proposed GMM-P was implemented in our battery swapping service concept. This paper generated the scenario with EVs and BSS inside the service area. The performance of GMM-P compared with the traditional random path, nearest job next (NJN), and k-mean with two and three clusters methods. According to the results, the average distance of GMM-P can achieve the shortest distance. Moreover, the GMM-P can reduce more than half of the distance from a traditional random method.