Distribution Network Load Forecasting Based On Active Load

With the development of smart grid construction,the distribution network becomes more intelligent, the interaction between the power grid and users becomes more obvious. Compared with traditional load, the active load like electric vehicles and interruptible load in the distribution network increased year by year. They can be controlled by the dispatching system and participate in the peak load regulation. They decrease the peak load of the power system. We need to consider the particularity of them when we carry out the load forecasting. So it is significant to study on the new load forecasting method under consideration of active load.This paper proposes a new load characteristic indicator of active load —— peak clipping power. Then the load curve of active load in the distribution network is acquired based on mathematical derivation and Monte Carlo sampling. Besides, an expression to calculate the peak clipping power of active load is proposed. This expression can provide a theoretical basis for the total and spatial load forecasting re-correction.This paper presents a new method of total load forecasting based on multiple data sources. This method has good compatibility and scalability, it uses a variety of data sources as input and take the various kinds of factors into consideration. On the basis of the total load forecasting results, this paper shows the corrected value under consideration of the impact of active load. The results show the expression to calculate the peak clipping power is applied in the total load forecasting. Then according to the three-layer mesh generation, this paper carries out the spatial load forecasting based on the improved trend prediction method. On the basis of the spatial load forecasting results, this paper shows the corrected value under consideration of the impact of active load. The results show the expression to calculate the peak clipping power is applied in the spatial load forecasting. So it provides an important theoretical and practical basis for the actual spatial load forecasting.Finally this paper uses the Python language to develop a load forecasting assistant decision-making platform for the traditional load forecasting. This platform can be used to complete the total and spatial load forecasting of a given area.