This paper introduces a novel weighted fusion methodology for grid-level short-term load forecasting that addresses the critical limitations of direct aggregation methods currently used by regional dispatch centers. Traditional approaches accumulate provincial forecasts without considering regional heterogeneity in load characteristics, data quality, and forecasting capabilities. Our methodology implements a comprehensive evaluation index system that quantifies forecast trustworthiness through three key dimensions: forecast reliability, provincial impact, and forecasting complexity. The core innovation lies in our principal component analysis (PCA)-based weighted aggregation mechanism that dynamically adjusts provincial weights according to their evaluated reliability, further enhancing through time-varying weights that adapt to changing load patterns throughout the day. Experimental validation across three representative seasonal periods (moderate temperature, high temperature, and winter conditions) substantiates that our weighted fusion approach consistently outperforms direct aggregation, achieving a 24.67% improvement in overall MAPE (from 3.09% to 2.33%). Performance gains are particularly significant during critical peak periods, with up to 62.6% error reduction under high-temperature conditions. The methodology verifies remarkable adaptability across different temporal scales, seasonal variations, and regional characteristics, consistently maintaining superior performance from ultra-short-term (1 h) to medium-term (168 h) forecasting horizons. Analysis of provincial weight dynamics reveals intelligent redistribution of weights across seasons, with summer months characterized by Jiangsu dominance (0.30–0.35) shifting to increased Anhui contribution (0.30–0.35) during winter. Our approach provides grid dispatch centers with a computationally efficient solution for enhancing the integration of heterogeneous forecasts from diverse regions, leveraging the complementary strengths of individual provincial systems while supporting safer and more economical power system operations without requiring modifications to existing forecasting infrastructure.

As renewable energy continues to be widely integrated, the energy structure is gradually transforming. The increasing grid connection of wind and photovoltaic power signifies a major shift in the energy mix. This change is particularly evident in heavy load areas at the regional grid and provincial dispatch levels, where uncertainties on both the supply and demand sides impact the daily operation of power systems. New dispatch strategies are urgently needed to address these uncertainties. This paper introduces a two-stage day-ahead and intra-day coordinated multi-level dispatch method that considers both the regional-level and provincial-level power systems, addressing supply-demand uncertainties from the perspective of regional grid-level and unmet load peak shaving. Unmet load refers to the load that cannot be met solely by the output of regional grid units. At the regional grid level, a unit dispatch model for unmet load peak shaving is developed. We introduce the concept of unmet load and, based on peak-valley weighting, propose a multi-province load peak shaving method, improving the approach to unmet load considerations. At the provincial level, a two-stage robust optimization dispatch model is constructed based on regional grid dispatch, and it is solved using the Karush-Kuhn-Tucker conditions and the Column-and-Constraint Generation (C&CG) algorithm. Finally, case study results validate the proposed model’s effectiveness, demonstrating its ability to provide an optimized coordinated grid-provincial dispatch strategy under supply-demand uncertainty.

As renewable energy continues to be widely integrated, the energy structure is gradually transforming. The increasing grid connection of wind and photovoltaic power signifies a major shift in the energy mix. This change is particularly evident in heavy load areas at the regional grid and provincial dispatch levels, where uncertainties on both the supply and demand sides impact the daily operation of power systems. New dispatch strategies are urgently needed to address these uncertainties. This paper introduces a two-stage day-ahead and intra-day coordinated multi-level dispatch method that considers both the regional-level and provincial-level power systems, addressing supply-demand uncertainties from the perspective of regional grid-level and unmet load peak shaving. Unmet load refers to the load that cannot be met solely by the output of regional grid units. At the regional grid level, a unit dispatch model for unmet load peak shaving is developed. We introduce the concept of unmet load and, based on peak-valley weighting, propose a multi-province load peak shaving method, improving the approach to unmet load considerations. At the provincial level, a two-stage robust optimization dispatch model is constructed based on regional grid dispatch, and it is solved using the Karush-Kuhn-Tucker (KKT) conditions and the Column-and-Constraint Generation (C&CG) algorithm. Finally, case study results validate the proposed model's effectiveness, demonstrating its ability to provide an optimized coordinated grid-provincial dispatch strategy under supply-demand uncertainty.

Grid-level dispatching is generally based on the accumulation of independent load forecasting data from provincial and municipal dispatch centers. However, the differences in economic development levels and the frequency of forecasting result updates among provinces and cities lead to certain limitations in the direct accumulation method, affecting the accuracy of the integrated forecasting results. To address this, this paper proposes a short-term load forecasting method for the power grid based on the i-Transformer model. First, the dataset is constructed through data preprocessing and feature engineering, followed by training and optimizing the model parameters. Further, considering the differences in forecasting results reported by provincial dispatch centers, principal component analysis is used to determine the weights of provinces and cities, thereby effectively integrating the forecasting data from different provinces and cities through weighting. The case study shows that the i-Transformer outperforms traditional statistical and machine learning algorithms on multiple evaluation metrics, and the integration method has considerable potential in handling multi-source heterogeneous data and improving forecasting accuracy. This paper provides a new means of load forecasting result integration for power grid dispatch centers, ensuring the safe, high-quality, and economical operation of the power system.