An approach of applying machine learning for battery optimization and range prediction for light duty commercial electrical truck

Abstract

Nowadays, Electric Vehicles are becoming quite popular, owing to their benefits like clean environment, free from noise, high power to weight ratio, adaptability for small power requirements, ease of portability, less staring time, higher efficiency with less chance of fluid leakages, less maintenance and lubrication consumption. In general, electric vehicles means, electric passenger vehicles. It is predicted that by the year 2030, commercial Electric Trucks (ET) are going to rule the world replacing the traditional trucks run by conventional fuels obtained from crude oil extracts viz. diesel, petrol and gasoline. These commercial vehicles have two types of duty cycles, light and heavy. The depletion in fossil fuels and the availability of conventional fuels, make fleet owners to move towards commercial electric trucks. But the alarming factor, which causes less attraction of ETs, is the capital cost. Though the maintenance cost is approximately zero, the initial cost, in turn the Target Cost of Ownership (TCO) makes them not to be so popular among the fleet management companies and owners. This research work, aims to analyse and provide a solution for reducing the TCO, based on the energy management system of the commercial ETs. Out of the total weight of the commercial ETS, 40% of the weight is contributed only through the energy management system, i.e. the battery system. . The vehicle parameters that affect the battery management system such as rolling resistance, tire pressure, payload etc. have been used as the inputs to predict the energy required for a particular trip schedule newline