基于WNN与FCM的电动汽车动态充电负荷预测方法

随着电动汽车动态无线充电(EV-DWC)技术的发展,针对目前EV-DWC负荷建模理论工作不全面的现状,以交通流量作为影响充电负荷的主要因素,以天气、典型日期、季节等因素为次要影响因素,根据路况建立负荷模型,通过电动汽车型号和状态的聚类不同对汽车分配不同的功率,完成动态充电负荷的建立。采用小波神经网络(WNN)对时序信息进行处理预测,再同误差反向传播神经网络(BPNN)相结合预测充电道路上的车流,短期车流预测精度为85%,用模糊C聚类(FCM)算法对电动汽车的充电类型以及该类型所对应的充电功率进行划分,将进入充电道路的电动汽车分为7种类型。根据各种充电类型分配相应的充电功率,完成日负荷建模。

Dynamic charging load prediction method of electric vehicle based on wavelet neural network and FCM

With the development of the electric vehicle dynamic wireless charging technology, Aiming at the current incomplete theoretical work of dynamic wireless charging modeling, this paper establishes dynamic charging load model according to road conditions, and assigns different power to different vehicles through clustering of EV models and states, so as to complete the establishment of dynamic charging load. Wavelet neural network is used to process and predict the timing sequence information, and then combined with back propagation neural network to predict the traffic flow on the charging road, and the short-term traffic flow forecasting accuracy 85%, with the Fuzzy c-means clustering algorithm to electric vehicle charging type and the type of the charging power division. There are seven types of electric cars entering charging roads. Fuzzy C-means clustering algorithm is used to divide the charging type of EV and the charging power corresponding to the type, and the EV entering the charging road was divided into 7 types. The corresponding charging power is allocated according to various charging types to complete the daily load modeling.