基于MATLAB的短期电力实用负荷预测模型的研究

Research on Short-term Electric Power Load Forecasting Model Based on MATLAB

为了解决电网规划决策中电力负荷预测问题,针对短期电力负荷预测系统来对系统进行预测。通过BP神经网络算法,运用MATLAB程序实现BP神经网络的训练过程,构建一个简单的短期电力负荷预测系统。通过对比,建立的神经网络系统的结构为24—9—24,即输入层的节点数为24、隐含层为9、输出层为24。通过对预测值与实测值的对比,可以得到误差基本保持在3%以内。可以通过BP神经网络对电力负荷进行预测。

In order to solve the problem of power load forecasting in power grid planning and decision-making, in this paper, the short-term power load forecasting system is used to predict the system. Through the BP neural network algorithm, the MATLAB program is used to achieve BP neural network training process and to build a simple short-term power load forecasting system. By comparison, the established structure of the neural network system is 24-9-24, that is, the number of nodes in the input layer is 24, in the hidden layer is 9 and in the output layer is 24. By comparing the predicted value with the measured value, the error can be kept within 3%. The power load can be predicted by BP neural network.