基于分层模糊神经网络的边缘侧光伏发电能量预测

Energy Prediction of Edge-side Photovoltaic Power Generation Based on Hierarchical Fuzzy Neural Network

以分布式光伏为代表的新能源广泛接入,带来业务实时响应、台区内智能分析处理等新需求,目前关于分布式新能源消纳本地自治的解决方案还鲜见报道,新能源消纳的关键在于光伏发电的准确预测与边缘物联装置研制。基于电网调度物联网,面向台区自治快速响应的需求,提出一种结合台区能量路由器的新能源消纳本地自治解决方案,研制面向分布式新能源消纳的能源控制器,并基于分层模糊神经网络算法对光伏发电的能量进行预测。研究表明,所研制的能源控制器以及提出的分层模糊神经网络模型在边缘侧台区自治中具有一定的效果和优势。

The wide access of new energy,represented by distributed photovoltaic,has brought about new demands such as real-time business responsiveness intelligent analysis and processing within the substations.The issue of local autonomy for distributed new energy consumption is rarely addressed in the literature.The key to optimizing new energy consumption lies in accurate prediction of photovoltaic power generation and rapid development of edge devices for Internet of Things.Based on the power grid dispatching Internet of Things,in this paper we propose a local autonomy solution for new energy consumption combined with the energy router of the substation.An energy controller is designed for distributed new energy consumption,and the energy produced by photovoltaic power generation is predicted based on hierarchical fuzzy neural network algorithm.The research results demonstrate that the energy controller developed in this paper,along with the hierarchical fuzzy neural network model,exhibit certain effects and advantages in enhancing the autonomy at edge-side stations.