A Novel of Adaptive Control Strategy for Multi-type Electric Vehicle Batteries

The development of electric vehicles (EVs) in the world is proven by the large number of EV users in the world. The main source of EVs is batteries of various types, so the Public Electric Vehicle Charging Stations (PEVCS) must be able to charge all types of batteries. The current problem is that PEVCS is unable to distinguish the detailed specifications of each type of battery, so the electric charging carried out on the battery is not optimal, thus reducing the battery life. Therefore, an adaptive PEVCS is needed to recognize each battery's detailed specifications. In this research, an adaptive control strategy that focuses on rapidly scanning the battery to detect its kind is used when the battery is connected to the PEVCS. In order to perform fast scanning of the battery, electric power is injected in a brief period of time, forming a characteristic pattern of current changes against voltage changes in each type of battery. The resulting data is then compared with four characteristic data of battery types that have been studied under various conditions using the Artificial Neural Network (ANN) algorithm, producing a precise and accurate detection. Following each battery type's preset point, the battery charging procedure is then completed using a consistent current and voltage. This research focuses on the detection of lithium-ion (Li-ion), Sealed lead acid (SLA), Nickel-Cadmium (NiCd), and Nickel Metal Hydride (NiMH) batteries, It is proven that ANN can detect four types of batteries accurately with a detection accuracy rate of 99% under various conditions and can perform optimal charging according to battery specifications.