Edge computing oriented virtual optical network mapping scheme based on fragmentation prediction

As edge computing services soar,the problem of resource fragmentation situation is greatly worsened in elastic optical networks(EON).Aimed to solve this problem,this article proposes the fragmentation prediction model that makes full use of the gate recurrent unit(GRU)algorithm.Based on the fragmentation prediction model,one virtual optical network mapping scheme is presented for edge computing driven EON.With the minimum of fragmentation degree all over the whole EON,the virtual network mapping can be successively conducted.Test results show that the proposed approach can reduce blocking rate,and the supporting ability for virtual optical network services is greatly improved.

基于近红外吸收光谱技术的高精度CO_(2)检测系统的研制

为了准确测量地震断裂带溢出的痕量CO_(2)气体浓度,文中采用可调谐半导体激光吸收光谱(TDLAS)技术,选取波数4978.202 cm^(-1)作为CO_(2)检测系统的吸收谱线,采用有效光程为40 m的多通池,以STM32作为主控和数据处理核心器件,研制了高精度CO_(2)检测系统。针对系统中的探测器噪声与光学干涉条纹噪声,利用卡尔曼-小波分析算法滤波提升系统性能。实验表明,与滤波前相比,系统在50 ppmv CO_(2)浓度下的二次谐波信噪比提升了2.06倍。在不同CO_(2)浓度下(50、300、1000、4000、8000 ppmv),系统误差为2.57%~2.66%。系统测量4000 ppmv浓度下的CO_(2)时检测精密度达到20.9 ppmv。利用Allan方差分析得出,积分时间在约61 s时对应的最低探测下限(MDL)为5.2 ppmv,实现了对CO_(2)气体的高精度测量。结果表明,所设计的高精度CO_(2)系统可以在气体检测领域为预测地震前兆提供良好前景。

采用线性自抗扰技术的高精度温度控制系统研制

随着光电检测技术的发展,红外气体检测技术在诸多领域有着广泛的应用。温度对于气体浓度及同位素丰度检测有着重要影响,采用比例-积分-微分(PID)控制算法的传统温度控制系统存在超调、响应时间慢和精度低的缺点。针对上述问题,首先使用COMSOL软件进行有限元分析,确定加热结构;然后以STM32单片机作为主控器件,通过16位AD芯片LTC1864进行实时温度数据采集;最后采用线性自抗扰算法(LADRC)算法调节PWM波,实现控制半导体制冷器(TEC)对系统温度的高精度实时动态调节。在19.8℃的环境温度下,进行目标温度为32℃的温控实验。结果表明,采用LADRC算法的温度控制系统在稳定工作时,温度波动标准差为0.035 7℃,相比于采用PID算法的温度控制系统,具有无超调、响应时间快和高精度的优点。

数字孪生辅助的智能楼宇多模态通信资源管理方法

多模态通信网络为智能楼宇能源调控数据的采集、传输、处理以及能源调控模型训练提供了通信支撑。数字孪生可以提供计算资源、信道特性等状态估计,辅助多模态通信资源管理优化,提高能源调控模型训练精度。然而,数字孪生辅助的智能楼宇多模态通信资源管理面临能源调控模型训练误差大、多时间尺度资源分配耦合、模型训练精度提高与能耗优化相互矛盾等挑战。针对上述挑战,提出基于数字孪生和经验匹配学习的多时间尺度通信资源管理优化算法,通过联合优化大时间尺度网关选择和小时间尺度信道分配与功率控制,最小化全局模型损失函数和能耗加权和。仿真结果表明,所提算法可以提高全局模型损失函数和能耗加权和性能,保障智能楼宇能源精准调控需求,促进智能楼宇能源调控低碳运行。

数字孪生辅助的智慧园区物联终端接入调度方法

数字孪生通过辅助智慧园区构建能源管理模型实现园区智能能源管理。本文面向数字孪生辅助的智慧园区能源管理场景,基于联邦学习的智慧园区数字孪生框架,构建了园区能源管理模型,在此基础上,提出了一种基于数字孪生和深度Q网络的终端接入调度算法。首先,构建了以能源管理模型损失函数最小化为优化目标,长期接入优先级为约束条件的接入调度优化数学模型;其次,基于虚拟队列将长期接入优先级约束和短期接入调度优化解耦,并结合数字孪生和深度Q网络实现优化问题求解;最后,通过仿真分析验证了所提算法的有效性。

回龙抽水蓄能电站发电电动机改造关键技术分析

抽水蓄能机组在高水头、高转速,双转向、频繁启停等运行条件下,逐渐暴露出一系列问题.近年来,国内抽水蓄能电站共计发生严重及以上缺陷近千个,以发电机和水轮机设备居多,其中发电机方面普遍存在推力轴承瓦温高、下导轴承甩油等缺陷,属于家族性问题.本文简要介绍了回龙抽水蓄能电站发电电动机在改造过程中遇到的问题,并对优化改造部分进行了分析总结,以期对国内外同类电站提供有益借鉴.

面向业务可靠承载的电力弹性光网络自主协同决策

针对电力通信骨干网面临频谱资源受限、调配灵活度差等挑战,提出一种面向业务可靠承载的电力弹性光网络自主协同决策方法。首先,建立长时业务可靠性约束,并通过虚拟队列漂移实现队列可靠性约束偏差感知;然后,将电力弹性光网络中业务与资源协同问题建模为业务请求、路径、频隙之间的三维匹配问题,并通过路径与频隙的聚合使其降维为一对一匹配问题;最后,引入升价匹配算法解决多业务请求对网络资源的竞争。仿真结果表明,相较于基于最小化队列漂移加惩罚的路由与频隙分配(MDPP-RSA)算法和基于首尾精确适配的路由与频隙分配(FLEF-RSA)算法,该文所提算法可有效提升电力业务可靠承载能力和网络资源利用率,平均业务队列积压降低了51.87%、80.18%,频谱利用率提升了33.52%、54.41%,并可满足不同优先级业务的高可靠传输需求。

基于工业互联网的设备电能监控系统

针对制造业智能制造升级改造过程中,设备电能监控管理的需求,提出一种基于工业互联网理念的设备电能监控系统的架构体系,研究系统的网络架构以及平台架构,进行电能监测,系统功能、系统界面的设计与开发。该系统采用智能电表监测设备的电能参数,通过数据采集平台把电表的Modbus协议转换成MQTT协议,并上传到云平台。在某工厂实际场景下,进行系统的安装部署及应用,实现设备电能参数的实时采集、监控与统计分析,实现该工厂的设备、产线、车间用电的精确计量,提高了设备用电统计分析效率,有效地支撑了工厂电能管理、用电优化,提高机器生产效率。该系统具有建设成本低、容易实现、运行可靠、利于复制推广的优点,可以基于工业互联网平台积累数据,进一步挖掘分析,构建能耗优化模型,实现工厂电能的智能管理。

基于边缘计算的智能配用电综合接入及状态监测技术研究

随着电力物联网建设发展需求的增加,人们对接入网络的广泛互联、时延性能和计算能力提出了更高的要求。针对接入终端多样性和数据量大等特点,考虑在设备采集过程中多协议适配可信传输的问题,建立基于边缘计算的设备接入与集成技术,引入差分算法和路由算法,提出基于边缘计算的差分检测方法,构建智能配用电通信综合接入设备本地数据智能化处理与业务融合模型,并将接入模型进行了应用仿真验证。该系统实现了设备和主站云平台之间的数据高效稳定传输,与远程云计算方式对比,降低了数据实时传输量,提升了时延、能耗等性能,具有分布式、低延时、高效率以及缓解流量压力等特点,相较单纯的云计算更加高效,可为智能配电综合接入及状态监测提供技术基础。

Physical layer impairments tolerance based lightpath provision in software defined optical network

As all-optical networks grow with ever increasing ultra-high speed,the communication quality suffers seriously from physical layer impairments( PLIs). The same problem still exists in software defined optical network( SDON) controlled by OpenFlow. Aimed to solve this problem,a PLIs tolerance based lightpath provision scheme is proposed for OpenFlow controlled optical networks. This proposed approach not only takes the OSNR model to represent those linear PLIs factors,but also introduces those nonlinear factors into the OSNR model. Thus,the proposed scheme is able to cover most PLIs factors of each optical link and conduct optical lightpath provison with better communication quality. Moreover,PLIs tolerance model is also set up and considered in this work with some necessary extension to OpenFlow protocols to achieve better compatibility between physical layer impairments factors and various services connections. Simulation results show that the proposed scheme is able to get better performance in terms of packet loss rate and connection setup time.

Naive-LSTM enabled service identification of edge computing in power Internet of things

Great challenges and demands are presented by increasing edge computing services for current power Internet of things(Power IoT)to deal with the serious diversity and complexity of these services.To improve the matching degree between edge computing and complex services,the service identification function is necessary for Power IoT.In this paper,a naive long short-term memory(Naive-LSTM)based service identification scheme of edge computing devices in the Power IoT was proposed,where the Naive-LSTM model makes full use of the most simplified structure and conducts discretization of the long short-term memory(LSTM)model.Moreover,the Naive-LSTM based service identification scheme can generate the probability output result to determine the task schedule policy of Power IoT.After well learning operation,these Naive-LSTM classification engine modules in edge computing devices of Power IoT can perform service identification,by obtaining key characteristics from various service traffics.Testing results show that the Naive-LSTM based services identification scheme is feasible and efficient in improving the edge computing ability of the Power IoT.

Dynamic fragments awareness based virtual network mapping strategy of elastic optical networks

As virtual networks services emerge increasingly with higher diversification, the issue of spectrum fragments presents great challenge to the elastic optical networks(EON), especially under heaven services burdens. Aimed to solve this problem, this article proposes a dynamic fragments awareness based virtual network mapping(DFA-VNM) strategy of elastic optical network. In this proposed approach, the dynamic fragments awareness model of it is established, which takes available bandwidth demand and spectrum fragment degree into consideration. Moreover, the dynamic fragments awareness based virtual network mapping strategy makes full advantage of real-time fragments awareness result to conduct virtual network service mapping operation with less fragments and lower blocking rate. Testing results show that the proposed approach is able to improved services supporting ability of EON.

Traffic-load prediction based on echo state network improved by Bayesian theory in 10G-EPON

With the evolution of 10-gigabit Ethemet passive optical network （10G-EPON）, the traffic-load prediction ability is necessary to support soaring services traffic with diversified characteristics and requirements. As a strong candidate to be used for the traffic-load prediction, the echo state network （ESN） may face the pseudo-regression problem and need to be improved for the better traffic-load prediction. To overcome this problem, this paper proposes an ESN based traffic-load prediction scheme using Bayesian theory in 10G-EPON for future-proof. In this proposed approach, Bayesian probability is introduced into the ESN and is used to improve the performance of ESN. According to the architecture between optical line terminal （OLT） and optical network units （ONU） in 10G-EPON, an ESN based on the Bayesian theory （B-ESN） is realized and the B-ESN based traffic load prediction scheme is also developed in OLT. Experiment results show that the proposed scheme can greatly better the accuracy of traffic-load prediction with lower complex degree.

Naive-LSTM based services awareness of edge computing elastic optical networks

Great challenges and demands are presented by increasing edge computing services for current elastic optical networks(EONs)to deal with serious diversity and complexity of these services.To improve the match degree between edge computing and optical network,the services awareness function is necessary for EON.This article proposes a Naive long short-term memory(Naive-LSTM)based services awareness strategy of the EON,where the Naive-LSTM model makes full use of the most simplified structure and conducts discretization of the LSTM model.Moreover,the proposed algorithm can generate the probability output result to determine the quality of service(Qo S)policy of EONs.After well learning operation,these Naive-LSTM classification agents in edge nodes of EONs are able to perform services awareness by obtaining data traffic characteristics from services traffics.Test results show that the proposed approach is feasible and efficient to improve edge computing ability of EONs.

Hierarchy echo state network based service-awareness in 10G-EPON

Aimed to enhance the supporting ability for diversified services, this paper proposes a hierarchy echo state network（HESE） based service-awareness（SA）（HESN-SA） mechanism in 10 Gbit/s Ethernet passive optical network（10G-EPON）. In this HESN-SA, hierarchy architecture is adopted to realize echo state network（ESN） classification based SA. According to the network architecture of 10G-EPON, the parent-ESN（p-ESN） module works in the optical line terminal（OLT）, while the sub-ESN（s-ESN） module is embedded in optical network units（ONUs）. Thus, the p-ESN plays the main function of SA with a total view of this system, and s-ESN in each ONU conducts the SA function under the control of p-ESN. Thus, resources allocation and transport policy are both determined by the proposed mechanism through cooperation between OLT and ONUs. Simulation results show that the HESN-SA can improve the supporting ability for multiple services.