Patent Grant
9158870
The present invention relates to Smart Grid (SG) communication networks. More in particular it relates to the simulation of Smart Grid communication and evaluation of different SG communication options based on different constraints based on the simulation.
The Smart Grid (SG) communication network is expected to be robust enough to handle various applications with unique requirements in terms of frequency of packets, packet size, delay, etc. Large scale simulations of SG communications, if feasible, would be valuable to identify the robustness issues. In accordance with various aspect of the present invention a toolkit has been developed to evaluate a mix of Smart Grid applications under specific network, topology and geographical constraints using a discrete event simulator. In one embodiment of the present invention the discrete event simulator is the OPNET Modeler®. The OPNET Modeler® and related products are developed and marketed by OPNET Technologies, Inc. of Bethesda, Md. The methods provided herein apply in general to any discrete event simulator that models a Smart Grid communications network.
Current simulation models of SG communications that include large numbers of users require very long simulation times and do not consider all relevant constraints.
Accordingly, novel and improved methods and systems for large scale simulations of SG communications which capture environmental characteristics such us terrain profile and population density and that mimic realistic results are required.
Aspects of the present invention provide systems and methods to model a communication network that supports an electrical Smart Grid. Operating behavior of the Smart Grid generates communication traffic in the communication network. The communication network operates under at least one communication protocol. Based on operating conditions of the Smart Grid the model can generate a network performance scoring index. The scoring index, which is based on low level statistics such as communication delay and message reception rate, gives an indication of how successful the studied communication protocol was for supporting the studied application.
In accordance with another aspect of the present invention, a system is provided to model a communication system in an electrical utility in a geography containing a first plurality of electricity devices, each electricity device transmitting data over a communication channel during a transmission time to a first node. The method includes the steps of determining from the first plurality of electricity devices a second plurality of electricity devices that are characterized as each having similar transmission characteristics and a similar communication channel to the first node; grouping of the second plurality of electricity devices into a plurality of bins, including a first bin, each bin capturing different electricity devices; and simulating by a processor of data traffic generated by the electricity devices captured by the first bin. The requirements of similarity are described herein.
The method can also include applying simulated data traffic generated by the electricity devices captured by the first bin to create simulated data traffic generated by the electricity devices captured by the remaining bins in the second plurality of electricity devices.
In accordance with one aspect of the invention, the transmission times of the electrical devices in the second plurality are distributed uniformly.
In accordance with another aspect of the invention, the transmission channel is a wireless channel and the first node is a wireless transmission tower.
In accordance with another aspect of the invention, the transmission performance of the electricity devices in the second plurality of electricity devices relative to the first node is determined by the processor based on a topological map of a location of the second plurality of electrical devices.
In accordance with another aspect of the invention, the method is applied in a discrete event simulator. The discrete event simulator can be applied to analyze a configuration of a communication network in support of an electrical energy grid.
In accordance with another aspect of the invention, the processor simulates data traffic in the electrical utility in the geography covering an operational period of at least 8 hours.
In accordance with another aspect of the invention, the simulated data traffic includes data related to Smart Meters and at least one other Smart Grid application in the group of electric utility Smart Grid applications consisting of Delivery Management and Optimization, Demand Management and Optimization and Asset Management and Optimization.
A system corresponding to the method is also contemplated. Thus a system to model a communication system which transmits data related to a first plurality of electricity devices in an electric utility in a geography including a device that is a first communication node. The system includes a memory to store data including instructions; a processor to execute instructions to perform the steps of: entering a topographical model of the geography; determining from the first plurality of electricity devices a second plurality of electricity devices that are characterized as each having similar transmission characteristics and a similar communication channel to the first node; grouping of the second plurality of electricity devices into a plurality of bins, including a first bin, each bin capturing different electricity devices; and generating simulated data traffic generated by the electricity devices captured by the first bin.
The system can perform the methods as described above and in the following description of the invention.
New Smart Grid applications will be supported via the deployment of robust information and communication infrastructures, which will enable the exchange of large quantities of data and control commands between the Smart Field Devices (home electricity meter or pole-top devices such as reclosers, capacitor banks, switches, sectionalizers, etc.) and the utility's Control Center. Currently, utilities from all over the world are facing transformations in their infrastructures and are assessing which available technology will satisfy their requirements. The long-term success of their strategic objectives, such as improved efficiency, integration of renewable energies or increased consumer engagement will directly rely on the communication infrastructure that they deploy today.
Good decisions are not straightforward in the vast landscape of available communication technologies (e.g. RF-Mesh, cellular, WiMax, Power-line Communication (PLC), point to point RF, private Wi-Fi). The choice of an optimal technology depends on a large numbers of factors, such as types of applications deployed, existing technology infrastructure, and the geographical characteristics of the deployment region.
In order to address the aforementioned challenges and assist utility companies to individually migrate and modernize their communication infrastructures, the Smart Grid Communications Assessment Tool (SG-CAT) has been developed, which is capable of simulating a mix of Smart Grid applications under various geographical topologies and topographies, user orientations, and applications configurations. Throughout herein the above tool will be identified with the acronym SG-CAT.
SG-CAT has been designed to exploit a discrete event network simulator, which allows it to reproduce realistic scenarios and simulate complex Smart Grid applications using a broad range of wireless protocols and technologies, such as Long-Term Evolution (LTE), WiMax, RF-Mesh or Private Tower systems.
Smart Grid Application Library for a Discrete Event Simulator
The ultimate goal of any Smart Grid deployment is to reach a level of robustness, reliability and security that allows the full implementation of a plethora of Smart Grid applications with different requirements and characteristics. In order to accomplish this target, detailed knowledge about these applications and their associated traffic models is essential. The following Table I summarized the characteristics of a number of Smart Grid applications to date.
In the above TABLE 1 AMI means “Advanced metering infrastructure”; DR means “Demand Response”; FA means “Feeder Automation” and EV means “Electric Vehicle.” A different characterization of Smart Grid applications is in three groups of electric utility Smart Grid applications: (1) Delivery Management and Optimization, (2) Demand Management and Optimization and (3) Asset Management and Optimization. These applications in one embodiment of the present invention include: Advanced Meter Reading, Remote connect/disconnect, Outage detection/last gasp, Fault location, isolation and service restoration (FLISR), Feeder automation (NOP/load balancing), Volt/VAR (reactive power) optimization, Faulted Circuit Indicator (FCI) Telemetry, Voltage regulator bank control, Transformer monitoring, Substation Remote Terminal Unit (RTU) connectivity, Demand response—baselining, Demand response—load control.
The performance of these applications in real environments will vary based on the communication technologies used to deploy them, the geographic elements (terrain and land use types) of the location and the topological nature (number of devices and location) of the network. Which technologies would be able to optimally handle all the communication requirements? What is the necessary throughput? Can all these applications be implemented in the existing utility's networks? General application requirements such low latency and high bandwidth are not enough to answer such questions. The actual application definitions need to be considered such as the packet size, the packet generation timing, synchronization among devices, etc. Without considering these aspects, packet collisions or network bottlenecks can't be understood,
In order to answer such questions insight is needed from large scale simulations that will allow reproducing realistic deployments.
With this goal in mind, an entire library of Smart Grid applications has been developed in accordance with one or more aspects of the present invention and the corresponding traffic models using discrete event simulators such as the modeling tools available in OPNET®. These tools enable a detailed definition of the different tasks included in each application and all the phases for each task, as can be seen in
The following applications are used:
A. Advanced Metering Infrastructure (AMI)
AMI is the first block, on top of which utilities can develop an entire library of Smart Grid applications. It allows direct connection between the Utility Servers and the user home Meters. It includes capabilities such us remote measurement readings, remote management as for instance described in [1] G. Deconinck, “An evaluation of two-way communication means for advanced metering in Flanders (Belgium),” in Instrumentation and Measurement Technology Conference Proceedings, 2008. IMTC 2008. IEEE. May 2008. pp. 900-905 and remote reporting as for instance described in [2] D. Hart, “Using AMI to realize the Smart Grid.” in Power and Energy Society General Meeting—Conversion and Delivery of Electrical Energy in the 21st Century, 2008 IEEE, July 2008, pp. 1-2. The infrequent uplink transmission and the short length of the packets have been reflected in the definition of the AMI tasks and phases as applied.
B. Automated Demand Response (ADR)
While the main beneficiaries of AMI are the utilities, through remote measurement readings, in the case of ADR both the utility companies and the customers will be able to take profit from its implementation. The utilities will be able to avoid peak consumption (and its associated high costs due to peak generation of energy) and the customers can save money by shifting their demands over time according to the real-time pricing schedules broadcasted by the Utility as described for instance in [3] S. Valero. M. Ortiz. C. Senabre, C. Alvarez, F. Franco, and A. Gabaldon, “Methods for customer and demand response policies selection in new markets,” Generation, Transmission Distribution, IET, vol. I, no. 1. pp. 104-110, January 2007. In ADR, the packet transmission is more frequent and the response should be quick enough for granting the proper service operation. Included in the OPNET library provided in accordance with various aspects of the present invention are methods for PULL and PUSH the information with different frequencies and packet sizes.
C. Feeder Automation (FA)
Automated supervision and control of substations allows overall coordination in case of emergencies and optimizations of operating costs. The communication system is a vital part of the wide area power system relaying and coordination. Relays isolate local failures in generation, transmission and distribution so that they do not spread to other parts of the grid. Distribution feeder automation refers to substation equipment for the detection, location and isolation of faults and a means to restore power to undamaged sections of lines. This functionality is referred to as Fault Location, Isolation and Supply Restoration (FLISR). A typical feeder system includes a circuit breaker and at least three of the following types of switching devices along the line: reclosers, disconnect switches, sectionalizers, airbreak switches and fuses.
Distribution networks need to evolve and transform the static and conventional grids into dynamic and reliable smart grids. Optimizing the operation and maintenance and improving the overall coordination in case of emergencies are challenges that the new grid will face. In order to accomplish these goals, a robust and trustworthy communication system is needed. However, nowadays, the level of communication and automation along the feeder and the distribution substations is really basic and will not be able to handle the future smart grid applications in security and isolation. In one embodiment a system distributes logic amongst relays to implement a decentralized control system. In an emergency, the fault location, isolation and supply restoration should be done as fast as possible. Thus the communication technology should be fast enough for handling fast transmissions with extremely low delay, as for instance described in [4] A. Smit. “Distribution Feeder Automation using IEC61850 GOOSE Messaging over WIMAX Wireless Communications.” Several technologies (GPRS, PLC, WIMAX, etc) have been proposed but there is no silver bullet that covers all applications. Therefore, simulations of different scenarios and technologies need to be performed.
D. Electric Vehicle Charging (EV)
Electric Vehicles will cause a significant growth in the energy demand in the upcoming years as described in [5] “United States Department of Energy, Energy Information Administration, 2011 Energy Outlook.” This and other studies show that Electric Vehicles will have a penetration of 5% of the market by 2020 and they will continue to grow even faster over the following years. This means that millions of vehicles need to be integrated into the power supply infrastructure. A large fleet of EVs needs to be managed in an intelligent way in order to optimize and control the charging of their batteries without generating uncontrollable load peaks. During the charging process, car chargers will need to communicate with the utility servers in a fast, secure and cost-efficient way. A realistic simulation of the different proposed standards (SIP, IEC61850, etc) is necessary for a realistic evaluation of the communication requirements. EV application presents different real time characteristics depending on the scenario: roadside chargers need to finish the process as soon as possible, while in-house chargers can schedule the charging for a longer period of time.
E. Others
As the grid gets more and more intelligent, the number of possible applications will increase. New applications, such us Mobile Workflow Management, Renewable Sources Monitoring, etc, will require more demanding channels with higher throughput and lower latency, specially for emergency situations and video surveillance applications. The herein provided library will allow the testing of such applications in realistic environments for evaluating the future performance of the Smart Grid communication networks.
Challenges of Implementing Smart-Grid Applications for Discrete Event Simulation
In order to model and evaluate the different communication capabilities of a given Smart Grid deployment and guide Smart Grid application implementations, the Smart Grid Communications Assessment Tool (SG-CAT) has been developed.
OPNET® was chosen as a basis for the herein provided communication simulation tool due to its powerful and high-fidelity simulation and modeling capabilities. OPNET® allows to creating a complete library of applications, to evaluate their behavior with different communication technologies, and perform simulations with realistic terrain. Although the OPNET Modeler® is powerful, modeling an entire network with thousands of wireless nodes, e.g. LTE or ZigBee, is still a challenging task because of computational limitations with scale-up. For wireless nodes, OPNET Modeler® creates message passing pipelines between each pair of nodes that represents the wireless medium from every user's perspective. As a result, the number of pipelines increases quadratically with the number of users, as shown in
In accordance with the discussion above, the number of message passing pipelines and the execution time are correlated.
Another challenge is due to the interest in studying RF Mesh technology in the Smart Grid. The OPNET library contains the Zigbee RF Mesh model. However, unlike many of the other models available in the library, the Zigbee application process does not make use of OPNET's standard application model object. Thus, it is currently impossible to use the Smart Grid application library for Zigbee studies, since the library has been developed within this framework. Presently, the Zigbee application process only allows nodes to behave as single traffic generators.
Although results will not be provided, it should be mentioned that a similar LTE deployment takes orders of magnitude longer to simulate. Additionally, the built-in efficiency modes (i.e. physical layer disabled) cannot be used to speed up simulation time, since one interest is in observing physical layer statistics and the effect of terrain and topology configurations.
It is also expected that a large scale power line communication or PLC network will be significantly faster. However, at present, PLCs are not a viable option as Medium Voltage/Low Voltage (MV/LV) transformers are known to kill PLC signals as described in [6] S. Galli, A. Scaglione and Z. Wang, “For the grid and through the grid: The role of power line communications in the smart grid,” CoRR, vol. abs/1010.1973, 2010. Until research allows communications through them, wireless alternatives seem to be the best option for Smart Grid communication. However, non-wireless communication technologies, including PLC networks are specifically included herein as an aspect of the present invention.
Approaches to Address Simulation Challenges
In order to accomplish realistic wireless simulations for a Smart Grid network with a large number of nodes, the challenges discussed above must be addressed sufficiently and effectively. The first step in the process of solving these challenges is to clearly identify an objective of the simulations in order to redefine them to a perspective of interest. This principle will be used below.
Scale-Up Challenges
The first task in determining how to address the issue of large scale simulation is identifying the minimum number of Meters that is needed to be studied. That is, ideally it would be desirable to simulate the entire service area of a utility (potentially having more than 100K Meters). However, considering the network in a hierarchical fashion, it is realized that it is made up of multiple small sub areas, namely individual cells containing a single takeout tower. The individual cells are interconnected via a backhaul to the central location and from a cell perspective behave independently, which indicates that it may be sufficient to study the cells sequentially.
It is not uncommon to find single cells that enclose about 1000-5000 Meters, which provides a huge scale-up advantage over the initial problem of 100K Meters with multiple cells. However, even this scale-up factor may not be sufficient enough as
If one wants to study overall throughput constraints in a communication network and reduce the number of simulation nodes, the most direct strategy is to group the nodes into “aggregated-nodes”, where each of the resultant nodes will transmit the sum of the individual nodes' traffic. This strategy can be optimal when one wants to study the capacity at the access point or in the backhaul area. Unfortunately, one would lose individual behavior and statistics, which are critical elements in evaluating Smart Grid applications.
Since one interest is in observing individual level statistics, such as end-to-end delay, scale up techniques must preserve a certain level of individualism. Although it seems like a tough task, certain assumptions and constraints imposed by the Smart Grid network help to simplify the problem.
Firstly, one can expect the network within a cell to have a somewhat balanced nature when the Meter density is high, since a typical suburban layout promotes a repetitive pattern. As the number of users increase, the probability that the resultant network will be scattered reduces. This observation is used to an advantage when creating the present consolidation rule in accordance with an aspect of the present invention, which is detailed further below.
Secondly, one benefits from the realization that for the non-real time Smart Grid applications, nodes are not required to transmit simultaneously (and will not, due to technological constraints). As a result, a pseudo-scheduler that organizes the transmission times to minimize channel access clashes is applicable. In real-time applications, such as Feeder Automation, the number of nodes is expected to be low and such applications are not as susceptible to scale-up issues as others.
The issue of large scale simulation for viewing individual level statistics is solved as an aspect of the present invention by introducing a method of consolidation that uses the principle of statistically independent time bins. This method can be explained as follows: assuming there are n users in the cell and all must transmit within t seconds, one can visually represent the transmission behavior of the network as a single bin of length t, capacity of n and average inter-arrival time of t/n (as shown in
1) The user transmission times must be distributed uniformly over time as otherwise the number of transmissions per bin (i.e. inter-arrival times throughout the entire simulation) would not be consistent. As the amount and the frequency of traffic generated in Smart Grid applications are rather small compared to other traditional applications, the network is expected to function near optimal with minimum number of clashes, if large transmit times (t) are assumed. Hence, this behavior can be approximated with a network scheduler that assigns arbitrary transmission times to each user with a uniform distribution.
2) The users that are split into different bins must be of a similar nature in order to exhibit an independent nature of each bin. If this criterion is not met, then each bin would behave differently and must be simulated separately. As typical Smart Grid deployments include several Meters located in close geographic proximity, the transmission characteristics of such Meters will be of a similar nature. In accordance with an aspect of the present invention a similarity includes a similarity in channel characteristics between the Meter and a tower. This may be expressed in a transmission loss or similar characteristics. In one embodiment of the present invention two Meter related channels are deemed to be similar if their path loss over a pre-defined frequency range does not differ more than 3 dB. In one embodiment of the present invention two Meter related channels are deemed to be similar if their path loss over a pre-defined frequency range does not differ more than 1 dB. In one embodiment of the present invention two Meter related channels are deemed to be similar if their path loss over a pre-defined frequency range does not differ more than 0.5 dB.
In one embodiment of the present invention one criterion for two Meters being similar is its distance to a tower. For instance, two Meters meet at least one criterion for similarity if their distance to a tower is the same within a 20% margin; two Meters meet at least one criterion for similarity if their distance to a tower is the same within a 10% margin; and two Meters meet at least one criterion for similarity if their distance to a tower is the same within a 5% margin.
In one embodiment of the present invention one criterion for two Meters being similar is its presence in a common neighborhood, for instance as defined by an area. For instance, two Meters meet at least one criterion for similarity if they both are located within an area of 25,000 m2. For instance, two Meters meet at least one criterion for similarity if they both are located within an area of 10,000 m2. For instance, two Meters meet at least one criterion for similarity if they both are located within an area of 1,000 m2.
When at least both of the criteria of transmission path and uniform distribution are met, one is able to create the so-called ghost bins as illustrated in
With Smart Grid deployment in mind,
In accordance with an aspect of the present invention, a determination of a similarity of communication channels of electricity devices such as smart meters related to a node such as a wireless tower is performed automatically or interactively by a computer. A topologic or topographic map is applied in accordance with an aspect of the present invention to determine a path loss for a wireless channel over a terrain with certain topological features. Accordingly, a computer can decide from a digitized map that meters located within a predefined area of a tower will have similar transmission channels to the tower. A user can assist in marking the digitized map for areas that would be considered as preferred areas for combining into bins, as shown for instance in
In accordance with an aspect of the present invention a computer searches one or more topological maps of predefined areas and selects meter locations with a similar loss to a tower as being enabled to be placed in one of a plurality of bins.
Application Challenges in Mesh Networks
As mentioned before, currently OPNET Modeler® does not include a wireless mesh technology, e.g. Zigbee, which is compatible with the application, task and profile objects available in the OPNET® palette. As a result, one is precluded from assigning multiple applications to a single node as well as designing new applications that require two way communications.
For this challenge, again advantage is taken of the non-real time behavior of Smart Grid applications by creating a super-application that aggregates the traffic of all the applications in one node with certain statistical distributions for the size and the inter-arrival time between packets. Knowledge of the application behavior allows to create packet distributions that can be applied to the mesh nodes.
Results
Testing of the consolidation method as described above is described next. For the simulations networks were created based on typical US suburban towns, which obey common Smart Grid deployment assumptions; hence, outlined conditions for the herein provided methods are met. Findings for the consolidation methods from the perspective of reception rate and end-to-end delay are also provided.
In
Hence, if reception rate and end-to-end delay are the statistics of interest, then the herein provided methods can be used safely without loss of accuracy of the results.
Further Communication Challenges
Terrain information is important in large scale studies, as hilly regions offer different challenges compared to flat regions. When considering the signal attenuation of a transmitter as a function of distance, many path loss models exist. Free space pathloss is not valid after some distance as foliage and obstacles begin to play a role. The Suburban pathloss model defined in V. Erceg, “An empirically based path loss model for wireless channels in suburban environments,” IEEE JSAC., vol. 17, no. 7 pp. 1205-1222, 1999 and which is incorporated herein by reference shows that wireless communication in an outdoor environment is affected by the amount of hilliness and tree density in the region.
Such an empirical model provides better results than free space on average, as it accounts for signal attenuators in the region. However, even such models rely on simple one-slope characterization of pathloss, which does not consider particular variations of the terrain. It is important to take the height of the antennas and the particular horizon seen by the antennas into account, since even connectivity might not be possible. Longley Rice described in A. G. Longley and P. Rice, “Prediction of tropospheric radio transmission loss over irregular terrain, a computer method-1968.” ESSA Technical Report ERL. 79-ITS 67 Institute for Telecommunication Sciences, Tech. Rep., 1968, which is incorporated herein by reference, is one such model that considers freespace propagation, ground reflection, terrain diffraction and for long antenna distances, scattering.
In an RF-Mesh deployment, the effective range of communication heavily depends on the absolute altitude and orientation of the transmitter and receiver. Access points and relays are typically mounted on pole tops and other high locations, whereas Smart Meters are more likely to be found closer to the ground. A Smart Meter herein is an electricity usage meter with two way communication capabilities and A Meter-to-AP hop would be the final hop, a Meter-to-relay hop would be used to overcome the coverage gaps, and a Meter-to-Meter hop would be used to form a long-range multi-hop mesh network. In a cellular deployment, the terrain profile between a Meter and base station directly impacts whether connectivity can be established or not. Unlike an RF-Mesh deployment, a cellular network cannot use intermediate devices to relay information and keep connectivity.
Regardless of technology in use, since all users share the same medium, the number of users in the network impacts the quality of service in terms of expected latency, delay jitter and achieved throughput. In extreme cases, it may even cause the network to collapse. Since each communication technology handles additional users differently, the network size should be considered carefully during the network planning phase.
The Smart Grid Communications Assessment Tool
Smart Grid applications and the different communication technologies supporting them present unique challenges in coverage, delay and reliability due to real-world terrain constraints, details of real-world deployment, user density and implementation details. For this reason a detailed study and simulation of the individual utility's network should be performed. In order to address these challenges the Smart Grid Communications Assessment Tool (SG-CAT) has been provided herein in accordance with various aspects of the present invention, which is capable of simulating a mix of Smart Grid applications under various geographical topologies, user orientations and application configurations. SG-CAT models and evaluates the communication capabilities of a given Smart Grid deployment, develops a list of communication options to consider and offers recommendations.
SC-CAT has been built on top of a discrete event simulator. In one embodiment of the present invention such simulator is the network simulator OPNET Modeler®, which is a powerful discrete event network simulator that provides high-fidelity modeling and detailed analysis for very large wired and wireless networks. The different OPNET Modeler® solutions enable the users to model various kinds of networks using a broad range of wireless protocols and technologies in realistic scenarios which take the effects of the terrain, the mobility of the nodes and multiple pathloss models into consideration.
SG-CAT enables the analysis of possible network topologies for diverse utility requirements and existing architectures, topology scale-up and analysis of end-to-end behavior of customized Smart Grid applications. Furthermore, it allows the user to tune up different network parameters and evaluate growth scenarios for performance studies. In addition, SG-CAT provides the capability of incorporating power systems components through co-simulation with an electric power network/transmission network simulator, such as the Siemens PSS®E simulation system, marketed by Siemens AG of Munich, Germany. Multiple co-simulation systems can run simultaneously and exchange information (such as boundary conditions, events, time steps, control signals, etc.) in a collaborative manner within the SG-CAT environment. This allows performing hybrid complex Smart Grid simulations that let the users understand the interactions at the design and analysis stages.
The following illustrates several aspects of the SG-CAT tool.
Simulation Setup
The following configuration was used to apply in the illustrative SG-CAT simulation. The RF-Mesh and cellular technologies are chosen as a focus of an analysis. The RF-Mesh module, based on IEEE's 802.15.4 g SUN work group specification, operates in the unlicensed 902-928 MHz ISM band with QAM64 modulation at 100 kbps. For testing the cellular technology an LTE network configured for 10 MHz Frequency Division Duplex was used.
Simulation Results
Terrain effects on RF-Mesh Range are illustrated in
The results indicate that when the receiver is placed at a low height, a communication range of about 1 km is feasible. However, the range extends to about 1.5 km when the receiver is situated at a higher location. These results depend on the terrain profile of the testing area and the effective ranges are expected to vary in other terrain settings. These results suggest that meter-to-meter hops are possible over the range of hundreds of meters and high altitude relays can be used to further extend range in case of coverage gaps.
Coverage: For testing coverage areas, a receiver was placed at a fixed location, representing the takeout point. The resulting connectivity of the networks (RF-Mesh and LTE) is observed in a range of 11 km in one direction. In
The terrain characteristics not only have an effect on connectivity, but also on the quality of service. An RF-Mesh experiences increased delay due to the necessity for additional hops; whereas in an LTE network, additional delay might come from bad channel conditions causing retransmissions and other link adaptation mechanisms inherent to LTE networks (i.e. a lowering of the Modulation and Coding Scheme (MCS) index). Note that these results only focus on the coverage in one direction, and the challenges are magnified when a 360 degree view is taken into consideration, since the terrain profile, and accordingly the obtained coverage will vary in each direction.
3) Network Size:
4) Multi-Application. The Smart Grid network is expected to handle multiple applications with diverse requirements (see for instance Table 1 above) concurrently. In one simulation, the focus is on two communication parameters: the transmission packet size and the transmission frequency in order to assess the RF-Mesh networks capability to satisfy different applications.
Additional details on applications in a Smart Grid and how these are applied in the simulation provided herein in accordance with various aspects of the present invention will be provided below.
One way to distinguish between the SG applications is by their communication characteristic. This is illustrated in
Further details on application characteristics and their requirements with regards to a communication network are illustrated in
Further transmission behavior of AMI as a one packet application is illustrated in
Transmission of DR pricing as a one packet application is illustrated in
Packet flow in DR Load profiling (Push) is illustrated in
Packet flow in DR Load shed request is illustrated in
One can thus create customized models generated by SG-CAT to test assumptions and conditions related to specific SG configurations, as is illustrated in
Accordingly, realistic, dynamic and configurable traffic models are provided that emulate or significantly approximates the real behavior of different Smart Grid applications by applying a definition of realistic protocol and transmission models for SG applications and by applying a computer system tools that convert the protocol and transmission models into traffic templates. As a further aspect of the present invention a template is provided for certain common applications, features or aspects of a SG communication network that can easily be adapted to specific local requirements or conditions. A user thus can select a template for a certain environment (rather than defining a configuration completely from scratch) and customize it for local conditions, such as topology or number of users and/or traffic requirements and the like.
How a study related to SG-CAT is conducted is described in the following.
A study may begin with SG-CAT automatically importing the topology of all devices (such as cap banks, meters, reclosers, FCIs, voltage regulators, substation RTUs, distribution transformers, line sensors, protection relays, etc.) within the utility's service area, along with terrain and elevation data. This is illustrated in
After importing terrain and asset information, SG-CAT simulates deployment of the most popular Smart Grid applications and communications technologies for each device in the utility's service area. A user can configure the tool based on desired communications specifications (power levels, frequencies, modulations, etc.) for instance via an input screen as illustrated in
Individual statistics, such as latency, packet loss and Signal-to-Noise Ratio (SNR), from every device are collected and can be easily viewed for every combination of communication technology and Smart Grid application included in the study. For instance. details about delays are shown in 4103 which is provided enlarged in
As a result of a simulation the expected differences between various approaches can be visualized as illustrated in 4103 in
The scoring formulas take raw statistics relevant to the application of study such as latency or reception rate and convert them to a score index that indicates how successful the technology was in implementing the application based on the application output requirements. In one embodiment of the present invention, the score is directly determined by the percentage of packets that is correctly received. For instance a correct reception rate of 87.3% will be scored as 8.73 on a scale from 0 to 10. Other scoring methods are possible and are fully contemplated. For instance, one may provide each application with a unique scoring formula particular to their output requirements.
Co-Simulation
The new grid will require automatic fault detection, voltage optimization, load management, continuity sensing, phase management and a host of other new techniques to manage the much more complex power generation system and energy network of the future. All these functions will overlay on communication infrastructure.
Advance large scale power grid modeling and simulation with connections to other critical infrastructure such as communication, transportation, oil and natural gas, water. Power grid models subsume load flow, stability, circuit problems, congestion, voltage stability, dynamic stability. They are distributed in nature, and use communication as enabling technology. Accordingly, the expected operations of the grid have to be supported by the communications network and the design of the grid will influence the configuration of the communications network.
So, in addition to a realistic network simulation of smart grid communications networks, an integration with simulations of dynamics in power flows, control, data management of utility/power networks with communications is provided as an aspect of the present invention. Such a co-simulation platform is illustrated in
The following steps for the co-simulation method in accordance with one or more aspects of the present invention are provided next:
Assess/design the communication network for a given requirement set of an electrical system.
Examples: given an electrical system, communication points (electrical components requiring inbound/outbound communication traffic) and traffic patterns, perform the following:
a. Network Planner: Design the network topology, i.e. what kind of system is needed? If mesh, where are the optimal locations for access points with wired backbone access? What should be the frequency/channel selections? etc. This functions as a planning tool to assist the engineers in designing their communication infrastructure for a given electrical system.
b. Next Generation Smart Grid Protocols and Algorithms: Design new algorithms/protocols which address the unique problems associated with the traffic generation characteristics of the electrical system. An electrical system, at first sight, resembles a very large sensor/control network since the majority of the electrical components are overseen to be deployed with some kind of a sensor which reports about the condition of the component, near future needs, capabilities, as well as control mechanisms that are set in response to sensor inputs. In such cases, the traffic generation rate per source is rather small, however, the challenge lies in handling this traffic in extremely large systems (i.e. the number of nodes producing the traffic is quite large). New protocols/algorithms can be designed and tested in a co-simulation environment in order to tackle these unique problems.
One such example is In-Network Data Aggregation. Instead of transmitting all the traffic coming from all the electrical components, the network can be smart enough to pre-process and aggregate the traffic as it is flowing through a network route. With a co-simulation environment, one can design and implement the new communication protocols/algorithms that solve the unique challenges in SG and simulate them realistically.
How problems that occur in the communication network will affect the overall functionality of the electrical network as any errors/disruptions occurring on the communication link will disable the electrical network components to communicate with each other. In order to measure/assess the level of these disruptions, perform the following:
a. Worst-case Scenario Simulations: The disruptions of the communication network can be introduced in a co-simulation in order to see the subsequent effects of these disruptions on the electrical network. For instance, if the communication between two communication nodes is lost for 1 hour and if the traffic flowing through these nodes is lost for that period, what happens to the electrical network? What is the cost of this disruption from the electrical systems' side? Does it cause the electrical system to dysfunction or is it a minor glitch?
b. Self-healing Wireless Networks: Self-healing capabilities of the communication network can be analyzed and new procedures for timely and efficient wireless restructuring mechanisms can be provided. In case of any disruptions (as above), a proper communication network should, in a very fast manner, recognize the setback, calculate the necessary measures to follow and act on the calculated decisions. For each different electrical network setup, the requirements for the self-healing capability might be different (for instance, some electrical systems may require extremely tight delay bounds, whereas the others do not). The co-simulation of the two different networks will enable to comparing the capabilities of a communication network through the requirements imposed by different electrical systems.
The methods as provided herein are, in one embodiment of the present invention, implemented on a system or a computer device. Thus, steps described herein are implemented on a processor, as shown in
Several protocols that can be applied in a communication network to support Smart Grid have been mentioned herein. Other communication protocols and/or technologies that can be applied include: VHF/UHF Digital Radio, 900 MHz DSS Radio, 900 MHz RF Mesh, 2.4 GHz RF Mesh, Cellular (LTE), Narrowband P2MP, Long Range Wi-Fi, Wi-Fi Mesh, Narrowband PLC, Broadband PLC, Unlicensed and Licensed Microwave and WiMAX. It is to be understood that use of other technologies and protocols, including Internet protocols is possible and fully contemplated.
The processor can be dedicated or application specific hardware or circuitry. However, the processor can also be a general CPU or any other computing device that can execute the instructions of 1802. Accordingly, the system as illustrated in
In one embodiment of the present invention, the system conducts a simulation or modeling of at least a communications network related or in support of an electricity utility which includes preferably Smart Grid elements in a geography with different characteristics, different assets and different usages and with different amounts and times of generation of data traffic associated with different Smart Grid or utility applications as provided herein and with different conditions of operations.
One condition of operations is the time of operations that is simulated. For instance, the simulation on the system may in one embodiment of the current invention simulate an operational period of at least 24 hours. This allows an analysis of the expected network performance during such a period. One may also simulate the network for a period that is shorter, for instance an operational period of 1 hour, or a period of 2 hours, or a period of 3 hours, or a period of at least 4 hours. During the simulation one can simulate the entire network and/or utility or just a part of it.
During the simulation one can also change the operational conditions of the environment, which affects the demand for electricity or which affects operational conditions such as power line interrupts and the like. For instance, one may make different assumptions about environment conditions such a temperature, which at high temperatures may drive up the demand for power for air-conditioning and may shift demand for other appliances.
One may also simulate power line interruptions and assess the consequences on the demands on the network over different periods.
In accordance with one or more aspects of the present invention a novel approach for Network Traffic Profile Aggregation for Efficient Discrete Event Smart Grid Network Simulations has been provided. Steps of various methods according to the novel approach have been implemented on a processor based system using controlled experimentation using datasets for a variety of network configurations. Thus, novel systems and methods and steps implementing the methods have been described and provided herein for Network Traffic Profile Aggregation for Efficient Discrete Event Smart Grid Network Simulations.
In accordance with one aspect of the present invention a system for co-simulation has been provided. In one embodiment of the present invention at least a first and a second system as illustrated in
A further explanation on consolidation is provided in the following.
One question to be answered is “what is meant by the same results?” This is illustrated in
Again the measure of interest (f) for the same node should be the same in network N and network N′. There are two similar versions of the problem (1) the average results are preserved and (2) the individual results are preserved.
The concept of binning is applied to obtain the reduction. Ideally, it is possible to simulate a part of the original network setup and get similar results provided that (1) the transmission times are distributed uniformly over time and (2) the nodes are of a similar nature in traffic generation characteristics. Smart Grid applications support both of these assumptions.
Assume that a network of N devices connects or transmits to a concentration point such as a wireless communication tower. The N devices all send packets of data to the tower. For instance all communication of the devices to the tower is one packet per period T per device. (In another example each communication has p packets or no more than p packets). Each packet has a fixed number of symbols or about a same number of symbols or bits. One assumption is that the transmission times of the packets are distributed uniformly over T and that each device has identical or close to identical transmission characteristics, including size of the packet and channel characteristics between the device and the concentration point (the tower).
The discrete event simulator has to simulate each device during its simulation. In case of the N devices over a period of time, a number of devices will generate traffic or packets and a complementary number will generate no traffic. In accordance with an aspect of the present invention devices are arranged in bins, in such a manner that N/k devices that generate traffic during a period T/k will be placed in one of k bins and the remaining N-N/k devices that do not transmit during this period, but do transmit in one of other periods during the period T are placed in different bins. Accordingly, the N devices will be arranged in k bins, each bin containing N/k devices, each bin transmitting the traffic of N/k devices in a period T/k which is due over a period T.
In accordance with an aspect of the present invention one bin bp, which represents N/k devices which transmit within a period T/k, for instance from tp:tp+1 is simulated in a discrete simulator to generate simulated traffic during the period tp:tp+1. The traffic simulated over that period is deemed to be representative for the traffic generated during the other bins in period T. One can then replicate the generated traffic generated during tp:tp+1 (k−1) times and use the (k−1) replicated data sets as simulation data for the remaining (k−1) bins in period T.
In accordance with a further aspect of the present invention one simulates a network over a period d*T covering for instance a period of one or more days. In that case, one simulates a single bin of the network for d different periods which are preferably distributed uniformly over d*T and replicates each of the simulated traffic of that bin (k−1) times to provide data that represents the complementary bins not covered by the simulated bin. In this manner a significant reduction in simulation time is obtained.
Since data collection of devices such as smart meters need not occur in real-time, one can actually force or program meters to transmit on pre-programmed times so that the transmission times of the meters are distributed uniformly.
One can re-state the problem as follows: How to decrease the number of users in a simulation environment without compromising the end-to-end characteristics? For instance one has currently N nodes and one wants just N/k nodes to simulate.
Solution 1 is Cluster-R: Randomly remove half of the users so that the resulting scenario has a smaller network size.
Solution II is cluster-G: Remove the nodes according to their geographical proximity, i.e. remove only if there are other nodes that can represent the geographical location.
It is to be understood that the present invention may be implemented in various forms of hardware, software, firmware, special purpose processors, or a combination thereof. In one embodiment, the present invention may be implemented in software as an application program tangibly embodied on a program storage device. The application program may be uploaded to, and executed by, a machine comprising any suitable architecture.
It is to be further understood that, because some of the constituent system components and method steps depicted in the accompanying figures may be implemented in software, the actual connections between the system components (or the process steps) may differ depending upon the manner in which the present invention is programmed. Given the teachings of the present invention provided herein, one of ordinary skill in the related art will be able to contemplate these and similar implementations or configurations of the present invention.
The following references provide background information generally related to the present invention and are hereby incorporated by reference: [1] G. Deconinck, “An evaluation of two-way communication means for advanced metering in Flanders (Belgium),” in Instrumentation and Measurement Technology Conference Proceedings, 2008. IMTC 2008. IEEE. May 2008. pp. 900-905; [2] D. Hart, “Using AMI to realize the Smart Grid.” in Power and Energy Society General Meeting—Conversion and Delivery of Electrical Energy in the 21st Century, 2008 IEEE, July 2008. pp. 1-2; [3] S. Valero. M. Ortiz. C. Senabre, C. Alvarez, F. Franco, and A. Gabaldon, “Methods for customer and demand response policies selection in new markets,” Generation, Transmission Distribution, IET, vol. I, no. 1. pp. 104-110, January 2007; [4] A. Smit. “Distribution Feeder Automation using IEC61850 GOOSE Messaging over WIMAX Wireless Communications,”; [5] “United States Department of Energy, Energy Information Administration, 2011 Energy Outlook.”; [6] S. Galli, A. Scaglione and Z. Wang, “For the grid and through the grid: The role of power line communications in the smart grid,” CoRR, vol. abs/1010.1973, 2010.
While there have been shown, described and pointed out fundamental novel features of the invention as applied to preferred embodiments thereof, it will be understood that various omissions and substitutions and changes in the form and details of the methods and systems illustrated and in its operation may be made by those skilled in the art without departing from the spirit of the invention. It is the intention, therefore, to be limited only as indicated by the scope of the claims.
The present application claims priority to and the benefit of U.S. Provisional Patent Application Ser. No. 61/527,207 filed on Aug. 25, 2011, U.S. Provisional Patent Application Ser. No. 61/527,211 filed on Aug. 25, 2011 and U.S. Provisional Patent Application Ser. No. 61/527,212 filed on Aug. 25, 2011, which are all three incorporated herein by reference.
| Number | Name | Date | Kind |
|---|---|---|---|
| 6278963 | Cohen | Aug 2001 | B1 |
| 7688723 | Lo et al. | Mar 2010 | B1 |
| 8010643 | Gray | Aug 2011 | B2 |
| 8195443 | Malloy | Jun 2012 | B2 |
| 20030128713 | Kang | Jul 2003 | A1 |
| 20040153303 | Tang et al. | Aug 2004 | A1 |
| 20090082888 | Johansen | Mar 2009 | A1 |
| 20090184689 | Kressner et al. | Jul 2009 | A1 |
| 20110282507 | Oudalov et al. | Nov 2011 | A1 |
| 20120054712 | Melvin et al. | Mar 2012 | A1 |
| 20120217975 | Drazan | Aug 2012 | A1 |
| 20120323381 | Yadav et al. | Dec 2012 | A1 |
| 20130036311 | Akyol et al. | Feb 2013 | A1 |
| 20130197835 | Jonsson et al. | Aug 2013 | A1 |
| Number | Date | Country |
|---|---|---|
| WO2010119290 | Oct 2010 | WO |
| Entry |
|---|
| Tom Verschueren, et al. “Architecture for Smart End-User Services in the Power Grid”. In Network Operations and Management Symposium Workshops (NOMS WKSPS), 2010 IEEE/IFIP, IEEE, Apr. 19, 2010 (pp. 316-322). |
| Number | Date | Country | |
|---|---|---|---|
| 20130060552 A1 | Mar 2013 | US |
| Number | Date | Country | |
|---|---|---|---|
| 61527207 | Aug 2011 | US |
