Chongqing Energy Internet Engineering Technology Research Center, Chongqing University of Technology, No. 69 Hongguang Avenue, Chongqing 400054, China.
College of Computer Science and Technology, Chongqing Technology and Business University, No. 19 Xuefu Avenue, Chongqing 400067, China.
Sensors (Basel). 2019 Dec 30;20(1):205. doi: 10.3390/s20010205.
The node energy consumption rate is not dynamically estimated in the online charging schemes of most wireless rechargeable sensor networks, and the charging response of the charging-needed node is fairly poor, which results in nodes easily generating energy holes. Aiming at this problem, an energy hole avoidance online charging scheme (EHAOCS) based on a radical basis function (RBF) neural network, named RBF-EHAOCS, is proposed. The scheme uses the RBF neural network to predict the dynamic energy consumption rate during the charging process, estimates the optimal threshold value of the node charging request on this basis, and then determines the next charging node per the selected conditions: the minimum energy hole rate and the shortest charging latency time. The simulation results show that the proposed method has a lower node energy hole rate and smaller charging node charging latency than two other existing online charging schemes.
在大多数无线可充电传感器网络的在线充电方案中，节点能耗率没有被动态估计，充电需求节点的充电响应相当差，这导致节点容易产生能量空洞。针对这个问题，提出了一种基于径向基函数（RBF）神经网络的能量空洞避免在线充电方案（EHAOCS），命名为 RBF-EHAOCS。该方案使用 RBF 神经网络预测充电过程中的动态能耗率，在此基础上估计节点充电请求的最优阈值，然后根据选择条件确定下一个充电节点：最小的能量空洞率和最短的充电节点充电延迟时间。仿真结果表明，与其他两种现有的在线充电方案相比，所提出的方法具有更低的节点能量空洞率和更小的充电节点充电延迟。
