无轨电车供电线网分路最大容量估算  

The Maximum Capacity Estimation of Trolleybus Grid

作　　者：田丽华[1] 姜久春[1] 张帝[1] 王伟[2] 

机构地区：[1]北京交通大学国家能源主动配电网技术研发中心,北京100044 [2]北京交通大学国家示范性软件学院,北京100044

出　　处：《电源学报》2015年第3期113-118,共6页Journal of Power Supply

摘　　要：目前世界环境、能源问题的日益突出,但随着电池技术的发展无轨电车迎来了发展新机遇。但是,无轨电车线网分路众多、分路结构不同和负荷分布情况不同等导致不同分路的输电线路损耗均不同。分路最多可容纳的车辆负荷即分路最大容量无法通过统一的潮流计算公式得到,进而线网的容量难以估算,故此采用黑箱模型将分路平均容量的评估转变为一个数学问题,同时用BP算法来实现。首先,分析了线网分路容量的限制因素及影响因素;然后,利用BP神经网络算法来估算分路的容量,并根据分路容量的限制因素通过DIg SILENT仿真软件运行分路模型来建立BP网络,进而获取BP网络训练样本;最后,根据分路容量的影响因素确定BP网络结构和参数并进行训练,保存训练好的BP模型,搭建预测界面并进行预测。预测实例分析结果证明模型预测结果理想。The world environment and energy problem become more and more serious. With the development of battery technology, the dual-source trolleybus gets an opportunity to develop. However, trolleybus grids are numerous, and the load distribution is different. The maximum vehicle load which is the maximum capacity of trolleybus grid can＇t be obtained by one flow formula calculating. This article uses BP model to change average maximum capacity assessment into a mathematical problems. Firstly, the limited factors and influence factors of grid capacity is analyzed. Secondly, using BP neural network algorithm estimates the capacity of shunt. According to the limited factors of grid capacity DIgSILENT is run to set the BP network and to obtain the samples. Eventually, according to the influence factors of grid capacity the structure and parameters of BP model are determined and trained. Saving the trained BP model builds an interface for prediction. The results prove that the BP model is ideal.

关 键 词：无轨电车 BP模型 最大容量 线网分路 

分 类 号：TM312[电气工程—电机]