基于改进差分进化算法优化极限学习机的短期负荷预测

负荷预测的精度直接关系到电网的供需平衡,影响着电网运营成本。针对传统预测方法精度不高的缺点,提出了一种改进的差分进化算法优化极限学习机的预测模型。由于极限学习机的输入权值和隐含层偏置对预测精度有很大影响,因而利用改进差分进化算法对极限学习机参数进行优化,提高了极限学习机的泛化能力和预测精度。研究结果表明:改进差分进化算法优化极限学习机对短期负荷预测精度有较高提升。

Short - term load prediction based on IDE - ELM

The forecast accuracy of the load directly relates the balance between supply and demandof the power grid and affects the operating cost of the power grid. Considering the poor accuracy of tra-ditional prediction method, a short-term load prediction model based on extreme learning machine( ELM) optimized by improved differential evolution algorithm ( IDE) is proposed. Since the inputweights and hidden layer bias of ELM have a great influence on the prediction accuracy, the IDE isused to optimize the parameters of ELM to improve the generalization ability of ELM and forecastingaccuracy. The results show that the IDE-ELM can obtain higher accuracy for short-term load pre-diction.