基于DPSO-BP网络模型的短期电力负荷预测

Short-term Power Load Forecast Based on DPSO-BP Network Mode

为准确、快速、高效地预测电网短期负荷,提出改进的粒子群算法(DPSO),并与BP算法相结合,形成改进的粒子群—BP(DPSO-BP)神经网络算法,用此算法训练神经网络,实现神经网络参数优化,得到基于DPSO-BP算法的神经网络模型。算例分析表明,与传统BP神经网络法和PSO-BP神经网络方法相比,该方法改善BP神经网络的泛化能力,预测精度高,收敛速度快,对电力系统短期负荷具有良好的预测能力。

In order to improve short-term power load forecast accurately, quickly and efficiently, this paper puts forward a developed DPSO-BP neural network algorithm after connecting the DPSO algorithm connecting with the BP algorithm. The neural network was trained by the DPSO-BP algorithm, so as to realize the optimization of parameters of neural network and result into the neural network mode based on the DPSO-BP algorithm. The example analysis shows that the DPSO-BP algorithm effectively improves the generalization capacity of the BP neural network, has higher forecast accuracy and quick convergence speed, when compared with the traditional methods of the BP neural network and the PSO-BP neural network. As a result, the mode has an excellent forecast capacity for the short-term load of power system.