Hanting Li

Electric vehicle charging behavior exhibits strong randomness, resulting in complex and uncertain factors influencing charging demand. To accurately characterize the multidimensional and dynamic variations of these factors, this research proposes a Bayesian network–least squares support vector machine (BN–LSSVM) ensemble model under Bayesian optimization (BO). This approach enables comprehensive multi-factor analysis while specifically highlighting the impact of time-of-use pricing on charging demand. First, considering the differences in charging demand among people on different dates and under different weather conditions, a multi-factor joint probability modeling—BN model was constructed to analyze the probability distribution parameters of charging demand under different combinations of factors. Subsequently, the LSSVM model is employed to integrate the probability characteristics of charging demand, constructing a nonlinear relationship between electricity prices and demand, and realizing the different effects of different strategies on charging demand. Furthermore, employing BO to optimize the regularization parameters and kernel parameters of LSSVM, thereby achieving hyperparameter adaptive global optimization. Finally, the article takes the historical data from the Honghai Technology Charging Station in Qilihe District, Lanzhou, over a 1-month period as a case study; the results show that the combined model BN–LSSVM–BO has the smallest mean absolute error of 1.66 and root mean square error of 2.76 compared to independent, unoptimized models, demonstrating better demand forecasting capabilities. The BN–LSSVM–BO peak-time floating strategy reduced peak-time demand by 29.4%, while the valley time discount increased valley time demand by 12.2%, verified the significant impact of time-of-use pricing strategies on charging demand.

Coal‐fired power plants remain a major source of global CO ₂ emissions, posing urgent challenges for climate mitigation. This study proposes an AI‐enabled optimal control strategy for solvent‐based post‐combustion CO ₂ capture (PCC), aiming to reduce energy consumption and enhance environmental performance. A deep learning model combining convolutional neural networks and long short‐term memory (CNN‐LSTM) is developed to predict multi‐step CO ₂ capture rates. This model is integrated with a multi‐objective grey wolf optimizer (MOGWO)‐based economic model predictive controller (EMPC), which dynamically adjusts key operating variables. Simulation results demonstrate that the proposed control scheme achieves a 12.4% reduction in capture cost and a 10.6% decrease in energy consumption, while maintaining a high CO ₂ capture efficiency of 95.3%. These improvements not only optimize process economics while significantly lowering the environmental impact of carbon capture systems. These improvements contribute directly to carbon mitigation, offering a promising pathway for cleaner and more sustainable operation of coal‐based power generation under dynamic industrial conditions.