改进的容积卡尔曼滤波(CKF)算法在短期负荷预测中的应用

An Improved CKF Algorithm for Short-Term Load Forecasting

电力系统短期负荷可视为非线性系统的输出,为了准确地预测电力系统短期负荷,引入了容积卡尔曼滤波(CKF)方法,并通过估计和修正模型中的状态转移矩阵,得到改进的自适应CKF算法,以适应非线性系统的时变性.用某地秋季22d的历史负荷数据建模对未来9d负荷进行预测,仿真结果证明改进的CKF算法预测电力系统短期负荷是实用而有效的.

The short term load of power system can be regarded as the output of nonlinear systems.For the sake of to increase the accuracy of short term load forecasting,cubature Kalman filter（CKF）is introduced into the field of short term load forecasting.Based on estimating and changing the state transition matrix,an improved CKF is proposed to adapt to environmental disturbances and time-varying processes;therefore better forecasting results could be acquired.Historical load data of 22 days in autumn are used to establish the load forecasting model to predict the load of future 9days.The results show that the improved CKF is practical and effective.