一种新的智能变电站网络流量预测方法研究

New Method of Smart Substation Network Traffic Prediction

随着智能变电站网络改造的逐步实施,智能变电站网络流程预测技术开始被业界重视起来,智能变电站网络流量一旦发生异常,将直接影响到继电保护装置动作的可靠性、快速性和灵敏性。论文首先将灰色理论和人工神经网络算法相结合,构建灰色神经网络模型并对其进行分析;然后在此基础上通过附加动量变学习速率法对灰色神经网络的权值更新策略进行改进,提出一种基于改进的灰色神经网络智能变电站网络流量预测模型;最后以智能变电站的站控层交换机网络流量数据为例,以采集的原始频率数据为基础进行仿真验证,实验表明,该模型预测精度高,收敛速度快,提高了智能变电站网络流量预测的准确性和快速性,保障电网安全运行。

With the intelligent network transformation substation gradual implementation of smart substation network flow forecasting technology began to pay attention to it by the industry, SmartSubstation network traffic in the event of ab- normal operation of protection devices directly affects the reliability, speed and agility. Gray theory and artificial neural net- work algorithm are combined to build and analyze gray neural network. Then on this basis, through additional momentum variable learning rate method, gray neural network weights update strategy is proposed to improve a kinds of gray neural net- work based on improved SmartSubstation network traffic prediction model. Finally, SmartSubstation station control layer switches network traffic data, for example, to collect the raw frequency data as the basis for simulation, experiments show that the model predicts high precision, fast convergence and improve the SmartSubstation network traffic prediction accuracy and rapidity, guarantee the safe operation of the grid.