Improving driver behavior is an important way to reduce vehicle energy consumption. Currently, many scholars have conducted extensive research on eco-driving behavior and proposed various eco-driving suggestions. However, there is a lack of evaluation methods specifically aimed at assessing the driving performance of pure electric bus drivers. In order to reduce the operating costs and energy consumption of electric buses, this study collected naturalistic driving data from electric buses. Firstly, the original data was pre-processed with unified sampling frequency, data cleaning, and parameter supplementation. Secondly, by selecting driver behavior characteristic parameters and vehicle operation parameters, the impact of bus driver driving behavior on energy consumption was analyzed. Based on the identified behavioral parameters that influence energy use, and taking each trip from the departure station to the terminal station as a unit, seven energy-related driving events were proposed: average start-up acceleration time, number of rapid accelerator pedal presses, duration of sustained high pedal opening, number of sudden accelerations, braking proportion during deceleration, duration of low-speed driving, and duration of economic speed driving. Afterwards, a multiple regression model for eco-driving level evaluation was established by analyzing the Pearson correlation coefficients between various driving event parameters and energy consumption per 100 kilometers, and the driving level of the driver during the journey was scored. Finally, based on the evaluation model, an eco-driving assistance feedback platform was built to help fleet managers better understand the eco-driving level of drivers. The results show that the proposed eco-driving level evaluation model for electric buses based on driving events has an accuracy of 93.52% and an average error of 6.48% in evaluating the eco-driving behavior. The model has a good effect on calculating eco-driving scores. The eco-driving assistance feedback platform can help fleet managers understand the operation status of buses and the eco-driving level of drivers, and help drivers understand their own driving situation.

To identify the driving activities of urban buses while avoiding privacy infringement on drivers and other road users caused by the use of on-board surveillance cameras, this study establishes a bus vehicle behavior prediction model that takes vehicle motion and driving operation data as inputs. First, experiments were carried out to collect the natural driving data of urban buses, and vehicle movement and driver behavior operation data were collected through the CAN protocol. Then, segments corresponding to station entry, station exit, intersections, turning and lane changing were selected. Based on Takens’ delay embedding method, phase space reconstruction was performed to map time-series data into a high-dimensional space to generate two-dimension recurrence plots. Afterwards, multi-channel stacking was applied to construct RGB images. To address the issue of class imbalance, Focal Loss function was adopted to enhance the model’s feature extraction capability for minority classes. On this basis, an E-bus vehicle behavior prediction model marked as E-VBPM was developed using the ConvNeXt network. The results indicate that E-VBPM achieves an accuracy of 84.62% in predicting 5 kinds of driving activities. As compared with the machine learning algorithm that uses time-series data as the input, the proposed model achieves an absolute increase in accuracy, precision, and recall by 6.79%, 10.98% and 8.86%, respectively. The results of this research provide support for electric bus on-board systems to identify the current operating modes and assist the driver in a safer and more intelligent way.