An improved WM method based on PSO for electric load forecasting

作者：Yang Xueming^*; Yuan Jiangye; Yuan Jinsha; Mao Huina

来源：Expert Systems with Applications, 2010, 37(12): 8036-8041.

DOI：10.1016/j.eswa.2010.05.085

• Particle swarm optimization (PSO)
• Electric load forecasting
• Fuzzy systems
• Wang-Mendel (WM) method

The fuzzy system is an important method for intelligent modelling of electric load forecasting, and how to enhance the learning and data mining ability of fuzzy system is crucial for its practical application and the improvement of the load-forecasting accuracy. In this study. a PSO-based improved Wang-Mendel (WM) method is proposed, which is a new combined modelling method based on fuzzy system and evolutionary algorithm. This method adopts a modified Particle swarm optimization (PSO) algorithm to optimize the fuzzy rule centroid of data covered area and thus obtains complete fuzzy rule set through extrapolating. The electric load-forecasting model based on this proposed method is described, and a case study on short-term load forecast illustrates that this method effectively enhances the forecast accuracy of WM method, has a fast convergence rate, and is independent of the forecasting objects.

• 出版日期2010-12
• 单位华北电力大学