EVALUATING ENERGY-EFFICIENT DRIVING PATTERNS OF ELECTRIC BUSES AT STOPS

Abstract

Eco-driving techniques can drastically lower the energy consumption and operating costs of electric buses, especially when approaching and departing stops that call for frequent acceleration and deceleration. In this Paper, we provide a methodology for evaluating the eco-driving capabilities of electric buses by gathering data from real-world scenarios and analysing driving behaviour. The collected data underwent pre-processing, which included data cleaning, parameter addition, and sample unification. By analysing a range of energy-related driving events, including acceleration patterns, braking behaviour, pedal operation, and the duration of economical speed, we were able to identify critical driving behaviour parameters that impact energy consumption. A multiple regression model and a machine learning recognition framework were developed to evaluate eco-driving efficiency and categorise driving behaviours as either eco-driving or non-eco-driving. Implementing eco-driving techniques, which enhanced energy recovery and decreased power consumption at bus stops, was made possible by optimising NSGA-II and driving modes. Significant energy savings, increased assessment and recognition accuracy, and robust support for fleet management and driver assistance systems were all demonstrated by the results. The proposed model promotes sustainable urban transportation by lowering operating costs, boosting energy efficiency, and improving intelligent electric bus driving management.