Search

• A deep learning approach to predicting the electric vehicle (EV) charging demand. • A real-world large EV trajectory dataset was used for model training and validation. • The deep learning approach outperformed traditional time series prediction models. • Input data structures, and sample sizes were not influential to prediction accuracy. Short-term prediction of the Electric Vehicle (EV) charging demand is of great importance to the operation of EV fleets and charging stations. This paper develops a Long Short-Term Memory (LSTM) neural network to predict the EV charging demand at the station level for the next few hours (e.g., 1–5 h), using a unique trajectory dataset containing over 76,000 private EVs in Beijing in January 2018. To explore the performance of the LSTM model, we set up four scenarios by 1) comparing LSTM against two typical time series prediction models, i.e., the Auto-Regressive Moving Average model (ARIMA), and the Multiple Layer Perceptron model (MLP), 2) and investigating how different input data structures, sample sizes, and time spans and intervals would influence model accuracy. The results suggest that the LSTM model outperformed the ARIMA, and MLP models, and their MAPE1 values are 6.83 %, 21.58 %, and 18.31 %, respectively. In addition, we find that the time span and interval tend to be more influential to the LSTM model's prediction accuracy than input data structures, and sample sizes. In general, the LSTM model with a shorter time span or interval (e.g., 1 h) would perform better. [ABSTRACT FROM AUTHOR]

• A 3-level fragment division method is proposed for actual vehicle data processing. • The influencing factors of electric vehicle energy consumption are discussed. • A novel data-driven energy consumption prediction framework is proposed. • The driving conditions prediction is integrated in the proposed framework. With increasing mass-adoption of electric vehicles, the energy consumption has become a key performance index to electric vehicle drivers, automakers and policy-makers. Accurate and real-time energy consumption prediction under real-world driving conditions is essential for alleviating the 'range anxiety' and can provide support for optimal battery sizing, energy-efficient route planning and charging infrastructures operation. In this paper, real-world driving data collected from fifty-five electric taxis in Beijing city are obtained and divided into three-level driving fragments. The influencing factors of energy consumption, including vehicle-, environment-, and driver-related factors, are extracted and studied. With the extracted key influencing factors, a novel machine learning-based energy consumption prediction framework integrated with driving condition prediction is proposed and used in actual energy consumption prediction. The real-world trip test results show that a root mean squared error of 0.159kWh (RMSE) and a mean absolute percentage error 12.68% (MAPE) are reached, the RMSE and the MAPE are respectively reduced by 32.05% and by 30.14% compared to the conventional method. [ABSTRACT FROM AUTHOR]