基于PSO-ELM模型的短期电力系统负荷预测

Short- term power load forecasting based on PSO -ELM model

电力作为国民生活的必需之一，是社会发展、经济建设的基础，而提前对电力负荷做出预判能够提高电网运行的效率和电力分配的合理性。针对提高电力负荷预测的精确度、降低预测误差，提出了一种基于粒子群优化的极限学习机（PSO—ELM）模型。通过粒子群算法对极限学习机的参数进行寻优，最后通过PSO—ELM模型和传统ELM模型预测结果的相对误差比对，改进模型将相对误差降低在1％左右，提升了学习速率和预测精度，为电网运行和电力分配提供了决策保障。

Power, as one of the necessities of people- life, is the foundation of social development and economic construction. Making predictions of power load in advance can improve the efficiency of the power grid operation and the rationality of electric power distribution. To improve the accuracy of the prediction and reduce the prediction error, we proposed a kind of extreme learning machine based on particle swarm optimization（PSO - ELM） model. By particle swarm optimization （PSO） algorithm for extreme learning machine parameters optimization, by comparison of the relative error of the prediction results of the PSO - ELM model and the traditional ELM model, the improved model would reduce the relative error by 1%, thus improving the learning rate and accuracy and providing a decision - making power grid operation and electric power distribution.