Power Management for Controlling Event Detection Probability of Supercapacitor Powered Sensor Networks
Power Management for Controlling Event Detection Probability of Supercapacitor Powered Sensor Networks
复制标题
DOI：
10.23919/acc.2018.8431137
复制
发表时间：
2018-06
期刊：
影响因子：
--
通讯作者：
Ruizhi Chai;Ying Zhang;Geng Sun;Hongsheng Li
中科院分区：
文献类型：
--
作者：
Ruizhi Chai;Ying Zhang;Geng Sun;Hongsheng Li
在本文中，超级电容器供电系统的电源管理方法是针对复杂的传感器网络系统开发的，其中传感器占地面积取决于超级电容器的可用能量。以雷达传感器网络为例，证明了传感器网络的事件检测概率可以解耦为各节点的服务质量。因此，维持事件检测概率的问题被表述为跟踪每个传感器节点的参考服务质量值的问题。在该问题的表述中，超级电容器模型和网内处理模型被用作优化问题的两个约束，其中超级电容器模型捕获自放电和电荷重新分配现象，以实现存储能量的全部潜力。采用模型预测控制来解决粒子群优化的优化问题，仿真结果表明，该方法能够跟踪所需的服务质量，同时满足系统约束并保证超级电容器的端电压在正常工作范围内。
In this paper, the power management method of supercapacitor powered systems is developed for a complex sensor network system wherein the sensor footprint depends on the available energy of supercapacitor. With a radar sensor network as an example, it is proved that the event detection probability of the sensor network can be decoupled as the quality of service of each node. Accordingly, the problem of maintaining the event detection probability is formulated as a problem of tracking a reference quality of service value for each sensor node. In this problem formulation, the supercapacitor model and the in-network processing model are used as two of the constraints of the optimization problem, wherein the supercapacitor model captures both the self-discharge and charge-redistribution phenomena to achieve the full potential of the stored energy. Model predictive control is employed to solve the optimization problem with particle swarm optimization, and the simulation results demonstrate that the developed method can track the required quality of service while satisfying the system constraints and ensuring the terminal voltage of the supercapacitor to be within the normal working range.