基于OPTICS聚类的电力通信网跨层保护系统设计

电力通信网跨层传输数据时,受到噪声数据影响,导致跨层数据传输耗时长、电流畸变率高。为了保证跨层通信安全,设计基于密度的聚类算法(Ordering Points To Identify the Clustering Structure,OPTICS)的电力通信网跨层保护系统。构建并联有源电力滤波器安全保护电路,保证三相四线制限幅后具有良好谐波补偿效果。通过跨层控制主动切换模块,实现目标通信链路下达指令的主动切换。对电力通信网跨层数据进行OPTICS聚类处理,结合Laplace机制添加对称指数分布噪声,将Laplace噪声添加到聚类簇中,输出添加噪声后结果存储在输出队列中,完成通信网跨层保护。由系统测试结果可知,该系统传输耗时少,且在10 s测试时间内跨层电流畸变率仅为21%,能够起到保护通信网跨层传输的作用。

基于多Agent强化学习的电力通信网跨层保护方法

针对当前方法存在数据传输成功率低、传输延迟时间长以及开销大等题,设计基于多Agent强化学习的电力通信网跨层保护方法。首先使用多Agent强化学习算法设定网络多路径协议,控制网络节点数据接收能力,然后构建网络跨层安全构架,设定相应网络模型作为网络跨层保护的基础,最后使用罚函数法对模型进行求解,保证函数解具有较高的可靠性,根据求解结果实现对网络跨层算法的优化,实现电力通信网跨层保护方法。实验结果可知,所提方法的收包率得到了明显提升,传输延迟时间缩短,开销低。

Wireless sensor networks in relic protection:deployment methodology and cross-layer design

This paper describes the deployment optimization technology and the cross-layer design of a surveil-lance WSN system applied in relic protection.Facing the typical technical challenges in the applicationcontext of relic protection,we firstly propose a deployment technology based on ant colony optimization al-gorithm(DT-ACO)to overcome the difficulties in communication connectivity and sensing coverage.Meanwhile,DT-ACO minimizes the overall cost of the system as much as possible.Secondly we proposea novel power-aware cross-layer scheme(PACS)to facilitate adjustable system lifetime and surveillanceaccuracy.The performance analysis shows that we achieve lower device cost,significant extension of thesystem lifetime and improvement on the data delivery rate compared with the traditional methods.