使用卡尔曼滤波修正蒙特卡洛算法的电动汽车充电负荷预测研究

Ｒesearch on Load Forecasting for Electric Vehicle Charging Using Monte Carlo Method Modified by Kalman Filter

摘要:

电动汽车将在未来迎来大规模的推广和应用，并对电网运行产生深远影响。而电动汽车充电负荷预测将是 分析电动汽车对电网运行产生影响的基础。传统的电动汽车负荷预测方法仅仅通过对电动汽车运行行为特征进行 分析，预测结果也仅仅是一种预期，即便考虑了多种典型场景，也难以准确描述随机性较大的电动汽车充电情况及充 电功率大小。通过对电动汽车充电设施设置充电功率采样观测点，使用测量值对蒙特卡洛预测结果进行卡尔曼滤波 修正，计算出较为符合系统状态的预测结果，并使用某行政区进行了算例演示，所提计算方法为研究电动汽车充电负 荷精确预测和有序充电提供了新的方法探究和思路。

Abstract:

Electric vehicles ( EV) are to be widely promoted and utilized，and will have great impact on power system． The load forecasting of electric vehicles is an element task for analyzing the impact． The existing Monte Carlo method is based on EV operation characteristics，whose output is merely an expectation value，far from accurate depicts on the actual situation． Within sampling real － time measurement，a Kalman filter is proposed to modify Monte Carlo method，and the method is applied in a district． This method can provide a new theoretical strategy and generate certain ideas for EV load forecasting and orderly charging．