Forecasting of medium and long-term maximum power load for offshore oilfields based on PCA-GRD-LWR model

The annual maximum load is an important theoretical basis for the reasonable allocation of power supply and determination of the installed capacity of the system. Accurate forecasting results can reduce the equipment investment and operating cost of the offshore platform. The annual maximum load is closely related to such influencing factors as the oilfield production and the mining stage. The internal connection and variation trend of each characteristic quantity which affect the power load demand and the maximum load were analyzed, and the principal component analysis (PCA) was used to process the characteristic quantities, transforming the characteristic quantities with strong correlation into unrelated principal components. The grey relational degree (GRD) between each principal component and the maximum load was calculated, which was used to distribute different weights to regression results. The locally weighted regression (LWR) forecasting model based on the grey relational degree was established, and the parameter in the LWR model was optimized by the particle swarm optimization (PSO). The effectiveness of the proposed method was verified by analyzing the historical data of an offshore oilfield. The results indicate that the prediction error of the medium and long-term load forecasting is less than 3%, and the maximum load forecasting results in next 10 years are presented. © 2023 University of Petroleum, China. All rights reserved.