The large-scale integration of electric vehicles (EVs) brings significant challenges to power grid operation due to the strong nonlinearity and non-stationarity of charging loads. To address the limitations of existing methods in noise suppression and temporal feature extraction, this paper proposes a short-term EV charging load forecasting model that integrates Sample Entropy (SE) with a TCN-BiGRU hybrid network. First, Isolation Forest and KNN are used for anomaly correction and missing value imputation. Then, CEEMDAN decomposes the load series, and SE is introduced to screen out high-noise components for sequence reconstruction. The reconstructed sequence is fed into the TCN-BiGRU framework, where TCN extracts multi-scale local features and BiGRU captures bidirectional long-term dependencies. Tested on a full-year real-world charging dataset, the proposed model outperforms GRU, BiGRU, and TCN-GRU across all seasons. Compared with GRU, the proposed model reduces MAE and RMSE by 57.8% and 61.3%, respectively, in winter, while improving R² from 0.827 to 0.978.

Accurate regional demand forecasting supports reliable operation of urban electric vehicle (EV) charging networks. However, public charging demand exhibits spatial heterogeneity, multi-scale periodicity, and uncertainty. This paper proposes Urban-CSTPNet, a multi-relational spatio-temporal probabilistic forecasting framework. Five semantically explicit graphs represent geographical adjacency, spatial distance, historical demand correlation, pricing-pattern similarity, and static regional attributes. Sample-level time-conditioned graph gating fuses these relations using historical demand states and calendar context. Independent recent, daily, and weekly branches capture short-term variation, daily repetition, and weekly regularity, and are combined through temporal gating. The model produces multiple conditional quantiles and applies horizon-specific conformalized quantile regression using an independent calibration set. Experiments on the Shenzhen UrbanEV dataset at 1, 3, 6, 12, and 24 h horizons achieve a mean absolute error (MAE) of 60.44, root mean squared error (RMSE) of 192.44, and Pinball Loss of 16.96. These values are 11.51%, 6.04%, and 9.50% lower than those of ST-MGF-Q. For calibrated 90% intervals, the prediction interval coverage probability (PICP), prediction interval normalized average width (PINAW), and Interval Score are 0.9024, 0.0175, and 349.60. Results confirm improved forecasting accuracy, probabilistic quality, and empirical interval reliability.