Dépôt numérique
RECHERCHER

Étude de performance et co-simulation d'infrastructures de communications intégrées optiques et sans fil pour les réseaux électriques intelligents.

Lévesque, Martin (2014). Étude de performance et co-simulation d'infrastructures de communications intégrées optiques et sans fil pour les réseaux électriques intelligents. Thèse. Québec, Université du Québec, Institut national de la recherche scientifique, Doctorat en télécommunications, 269 p.

[img]
Prévisualisation
PDF
Télécharger (11MB) | Prévisualisation

Résumé

Le réseau intelligent est une amélioration significative du système électrique du XXe siècle caractérisée essentiellement par une amélioration de l'efficacité énergétique par l'entremise de flux bidirectionnels d'information et d'énergie. Cette thèse porte sur l'étude multidisciplinaire des réseaux intelligents et des réseaux d'accès optiques-sans fil de prochaines générations, où la forte capacité des réseaux d'accès ne profite pas seulement au secteur des télécommunications, mais à l'ensemble des systèmes de la société, incluant les secteurs de l'énergie et du transport. En premier lieu, le premier modèle analytique des réseaux optiques-sans-fil émergents basés sur les technologies de réseaux optiques passifs et réseaux maillés IEEE 802.11n/ e/ ac est proposé, permettant de (i) évaluer/comparer des algorithmes de routage FiWi et (ii) trouver la borne théorique supérieure de débit de données maximal des capteurs smart grids sans affecter négativement les performances délai-débit de traffic régulier. Utilisant un modèle analytique FiWi des performances de la couche MAC et combinant des échecs probabilistes des fibres optiques et des stations de base, la disponibilité du réseau est modélisée et quantifiée, permettant ainsi de la comparer aux exigences d'applications smart grids. En second lieu, les applications de véhicules électriques et sources d'énergie renouvelables sont étudiées de façon multidisciplinaire en tenant compte à la fois des perspectives de technologies de l'information, de communications et du système électrique. Considérant la complexité de l'élaboration d'un banc d'essai smart grid, la co- et multi-simulation est une approche prometteuse permettant d'étudier l'ensemble des perspectives smart grids en intégrant plusieurs simulateurs. Plusieurs approches de multi-simulation sont possibles, avec ou sans mécanisme de synchronisation pour coupler les simulateurs. Pour montrer l'utilité de telles multi-simulations, les perspectives de communications et du système électrique sont couplées pour modéliser les performances d'un algorithme proposé (nommé Int.VGR) intégrant la coordination de véhicules électriques et les sources d'énergie renouvelables en échangeant des paquets via un réseau d'accès optique-sans fil. La perspective de technologies de l'information est ensuite ajoutée pour étudier une application de télécontrôle smart grid basée sur des configurations réelles d'un système de distribution en France, où la finalité de la multi-simulation est de permettre de tester et valider de bout-en-bout de nouveaux mécanismes smart grids. Par l'entremise d'un banc d'essai d'un système de distribution réel miniaturisé, un modèle de programmation dynamique de coordination de véhicule électrique permet de significativement améliorer les performances du réseau électrique tout en tenant compte des exigences des clients. Le modèle est démontré expérimentalement via un réseau hétérogène Ethernet couplé au réseau de distribution, où un modèle proposé de synchronisation de capteurs coordonnés permet d'améliorer la synchronisation d'arrivée des mesures comparativement à une approche utilisant des capteurs périodiques.

Abstract

The smart grid is an improved power system of the 20th century characterized essentially by a two-way flow of energy and information. In that sense, the smart grid is sometimes called Energy Internet, where actors and components exchange energy and data with each other, similarly to the Internet where components and actors exchange data over a shared infrastructure. Many wired and wireless communications technologies could be used for smart grid communications. According to IEEE P2030, the three main quality attributes for robust smart grid communications are latency, reliability, and quality of service. Communications latency between a given pair of source and destination nodes is expressed as the time when the message is generated and the time when it is received at the reception side. Reliability is formally defined as the ability to perform a certain task for given conditions during a certain period of time. According to Telecom Italia, the current trend in the access networks is the convergence of fiber-to- the-home (FTTH) for their durability, reliability, and energy efficiency. China Telecom deployed over 70 millions FTTH ports between 2004 and 2012. Furthermore, China Telecom predicts that the major upcoming trends are: (i) increase of fiber penetration (ii) more capacity and speed, from 1 Gbps passive optical networks (PONs) towards 40 Gbps in a single PON system and (iii) increase of the coverage by integrating wired, wireless, and mobile nodes. According to China Telecom, PONs will converge into fiber-wireless (FiWi) networks to combine their respective advantages. Optical networks offer huge capacity, immunity against electromagnetic interference, and wireless networks can be deployed quickly at a low cost and offer mobility. FiWi networks can also be enriched by integrating optical and wireless sensors. Such networks can use fiber Bragg grating optical-based sensors and wireless sensors compliant with IEEE 802.15.4 ZigBee. Integrating sensors allow to interact with real-world systems to monitor different parameters, including temperature, pressure, sound, etc. This is a great opportunity for the telecommunications and many economic sect ors , that is, sharing low-cost access networks for improved efficiency and sustainability. In the following, we quant if y the performance of emerging FiWi access networks and propose new mechanisms by means of probabilistic analyses. Such probabilistic analyses allow to quickly quantify the communications performance for large topologies. In smart grid solutions, innovative partnerships hold great promise to enable utilities and other players to share smart grid communications infrastructures investments by transitioning from the traditional vertical network integration model towards splitting the value chain into multi-tier business models. Current Gigabit-class PONs evolve into next-generation PONs, whereby high-speed 10+ Gb/s time division multiplexing (TDM) and long-reach wavelength-broadcasting/routing wavelength division multiplexing (WDM) PONs are a promising solution near-term candidates. On the other hand, next-generation wireless local area networks (WLANs) based on frame aggregation techniques will lever age physical layer enhancements, giving rise to Gigabit-class very high throughput (VHT) WLANs. We develop an analytical framework for evaluating the capacity and delay performance of a wide range of routing algorithms in converged FiWi broadband access networks based on different next-generation PONs and a Gigabit-class multi-radio multi-channel WLAN-mesh front-end. Our framework is very flexible and incorporates arbitrary frame size distributions, traffic matrices, optical/jwireless propagation delays, data rates, and fiber faults. We verify the accuracy of our probabilistic analysis by means of simulation for the wireless and wireless-optical-wireless operation modes of various FiWi network architectures under peer-to-peer, upstream, uniform, and nonuniform traffic scenarios. The results indicate that our proposed optimized FiWi routing algorithm (OFRA) outperforms minimum (wireless) hop and delay routing in terms of throughput for balanced and unbalanced traffic loads, at the expense of a slightly increased me an delay at small to medium traffic loads. This first analytical framework for FiWi networks does not take into account quality-of-service (QoS), which is quite important for smart grids. We next study the performance of multi-tier integrated FiWi smart grid communications infrastructures based on low-cost, simple, and reliable next-generation Ethernet passive optical network (EPON), emerging high-speed TDM and multi-channel WDM PONs of extended fiber reach and QoS enabled VHT WLANs in terms of capacity, latency, and reliability. By means of probabilistic analysis and verifying simulations we study the coexistence of human-to-human (H2H), e.g., triple-play voice, video, data, traffic and machine-to-machine (M2M) traffic originating from wireless sensors operating on a wide range of possible data rates, time scales, and duty cycles. Our analysis enables the quantification of the maximum achievable data rates of both event- and time-driven wireless sensors without violating given upper delay limits of H2H traffic. The obtained results can be used as a theoretical upper bound of coexisting M2M traffic for the design and realization of future yet unforeseen smart grid applications. We next study in more detail the availability, which is an important metric of the reliability quality attribute. Availability is qualitatively defined in the IEEE P2030 standard. However, the availability metric must be quantified in order to validate given smart grid application requirements. In recent related work, availability has been quantified for wireless and optical backhaul networks in terms of communications reachability, while in some other work availability was not formally defined in a fine-grained manner and was assumed to be known. We develop a novel probabilistic availability model for integrated PON and WiMAX networks in order to quantify this metric according to medium access control (MAC) protocol limits as well as fiber and base station failures. The obtained numeric results show interesting availability behaviors, including the impact on availability depending on the number of base stations. We also investigate optical traffic re-routing through WiMAX when fiber faults occur and show that there exists a maximum amount of re-routed traffic for maximizing availability. Furthermore, we investigate a scenario of real-world smart grid traffic configurations shared with regular traffic and find the maximum sensor data rate to meet the availability requirements. We next study the opportunity to efficiently use the WiMAX channel dedicated to the utility in Canada. By using experimental measurements of smart grid applications compliant with IEC 61850 in trace-driven WiMAX simulations, we show that the WiMAX MAC protocol efficiency decreases for an increasing number of stations. To avoid this shortcoming, a novel WiMAX MAC protocol is proposed and analyzed for smart grid applications, which uses lattice correlators to improve the throughput-delay performance significantly. For the considered configurations, the obtained maximum throughput of the proposed MAC protocol outperforms the current WiMAX MAC protocol by up to 41 %. Next, to provide insights into smart grid applications, we study the coordination of electric vehicles and renewable energy sources in a multidisciplinary manner by means of experimental demonstrations and co-simulation studies. To evaluate large-scale smart grid systems, co- and multi-simulation experiments can be modeled. Multiple simulation tools have been built and studied independently in the communications and power system perspectives of IEEE P2030 to study new smart grid applications. However, very few studies have been done on co-simulation by combining both perspectives in a multidiciplinary manner. The implementation details of a novel communications and power distribution network co-simulator based on OMNeT++ and OpenDSS are discussed. The novelty of the co-simulator is demonstrated by showing the impact of data rate-based and event-based sensors on reactive/uncoordinated control algorithms of plug-in electric vehicles (PEVs) to reduce critical voltage durations. PEVs have great potential of being the alternative for the next-generation of transportation. Uncoordinated PEY charging, however, may put a significant pressure on the distribution grid. Using a modified IEEE-13 Node distribution network of 342 residential customers, a converged fiber-wireless infrastructure based on EPON, WiMAX, wireless mesh network and sensor technologies to support coordinated charging of PEVs is proposed. To measure the performance of both the communications and power system perspectives of proactive scheduling algorithms and proposed reactive control protocols, our developed hybrid co-simulator based on OMNeT++ and OpenDSS is used. Co-simulation results show that the proposed low-cost communications infrastructure enables to efficiently schedule the charging of PEVs and quickly stabilize the voltage in a stress scenario. To extend this co-simulation study, a novel integrated vehicle-to-grid, grid-to-vehicle, and renewable energy sources (IntVGR) coordination algorithm is then proposed. Its focus is on providing a multidisciplinary study on implementing the proposed IntVGR scheme over a broadband fiber-wireless communications infrastructure by co-simulating both power and communications perspectives. For the power systems perspective, results show that the scheme is able to achieve a 21 % reduction in peak demand compared to uncontrolled charging, and a better performance in flattening the overall demand profile and maintaining network constraints in comparison to a benchmark scenario. The scheme also demonstrates to successfully coordinate PEVs to take maximum utilization of local renewable energy. For the communications perspective, the measured upstream traffic for executing the proposed IntVGR scheme on a residential area is found to be 1-2 Mbps with an end-to-end latency level of 1 ms. The scheme has also been validated from both perspectives in a sensitivity analysis with a higher PEY adoption rate. Note that this work focus on the communications and power system perspectives. A multi-simulation model is then proposed to measure the performance of all smart grid perspectives as defined in the IEEE P2030 standard. As a preliminary implementation, a novel information technology (IT) and communication multi-simulator is developed following a High Level Architecture (HLA). To illustrate the usefulness of such a multi-simulator, a case study of a distribution network operation application is presented using real-world topology configurations in France with realistic communication traffic based on mc 61850. The multi-simulator allows to quantify, in terms of communication delay and system reliability, the impact of aggregating all traffic on a low-capacity wireless link based on Digital Mobile Radio (DMR) when a Long Term Evolution (LTE) network failure occurs. The case study illustrates that such a multi-simulator can be used to experiment new smart grid mechanisms and verify their impact on all smart grid perspectives in an automated manner. More importantly, multi-simulation can prevent problems before modifying/upgrading a smart grid and thus potentially reduce costs to the utility. High penetration of renewable energy sources and electric vehicles (EVs) creates power imbalance and congestion in the existing power network and hence causes significant problems in the control and operation. Despite huge efforts from the electric utilities, governments, and researchers, smart grid is still at the developmental stage to address those issues. In this regard, a smart grid testbed (SGT) is desirable to develop, analyze, and demonstrate various novel smart grid solutions, namely demand response, real-time pricing, and congestion management. We first present a novel SGT developed in our laboratory by scaling a 250 kVA, 0.4 kV real low voltage distribution feeder down to 1 kVA, 0.22 kV. Information and communication technology (ICT) is integrated in the scaled-down network to establish real-time monitoring and control. The novelty of the developed testbed is demonstrated by optimizing coordinated EV charging realized through the synchronized exchange of monitoring and control packets via a heterogeneous Ethernet-based mesh network. The developed SGT is a step forward to (i) find practical problems and (ii) validate and experiment new smart grid mechanisms in realistic physical conditions.

Type de document: Thèse
Directeur de mémoire/thèse: Maier, Martin
Co-directeurs de mémoire/thèse: Joós, Géza
Mots-clés libres: efficacité énergétique; véhicules électriques; énergie renouvelables; réseaux intelligents; réseaux optiques-sans fil
Centre: Centre Énergie Matériaux Télécommunications
Date de dépôt: 23 sept. 2014 21:00
Dernière modification: 24 nov. 2015 16:09
URI: http://espace.inrs.ca/id/eprint/2374

Actions (Identification requise)

Modifier la notice Modifier la notice