Charging Time Prediction of Electric Vehicle Charging Pile via Momentum-incorporated Non-negative Latent-factorization-of-tensors with Swish-regularization

As electric vehicles (EVs) are wildly applied in our daily life, how to rationally deploy the public charging piles has become a thorny issue. Hence, it is necessary to comprehensively consider the utilized periods and installation location of charging piles to analyze the charging time of each charging pile. However, large-scale data collection and processing is inefficient and time-consuming. To address this issue, this paper constructs a Momentum-incorporated Non-negative Latent-factorization-of-tensors with Swish-regularization (MNLS) model. Its main ideas include: 1) incorporating a general momentum method into the model to achieve rapid convergence; and 2) utilizing a swish regularization scheme to address the over-fitting issue Experimental results on four real-world datasets demonstrate that the MNLS model achieves higher prediction accuracy compared to the state-of-the-art forecasting methods.