基于Spark和深度神经网络的短期电力负荷预测方法

Short-Term Power Load Forecasting Method Based on Spark and Deep Neural Network

针对单机计算资源不足和提高负荷预测精度,提出一种基于Spark和粒子群优化深度神经网络的短期负荷预测模型。通过引入Spark计算平台,将深度神经网络模型部署在平台上,对深度神经网络模型的网络结构和权重及阈值参数利用粒子群算法优化,再利用优化后的深度神经网络模型预测电力负荷。通过实验分析,结果表明提出的电力负荷预测方法不仅精度上还是运行效率上优于其他比较的负荷预测方法,而且并行性较好,运行效率优于单机电力负荷预测模型。

Aiming at the shortage of computing resources of single computer and the improvement of load forecasting accuracy,this paper proposes a short-term power load forecasting model based on Spark and particle swarm optimization(PSO)deep neural network.By introducing Spark computing platform,the deep neural network model is deployed on the platform,and the network structure,weight and threshold parameters of deep neural network model are optimized by using PSO,and then the optimized deep neural network model is used for power load forecasting.Experimental analyses have been carried out,and the results show that the proposed load forecasting method is not only superior to other load forecasting methods in accuracy and operation efficiency,but also has good parallelism,and the operation efficiency is better than that of single-computer load forecasting model.