Automation and Standardization Needs in Smart Grids

Application of the available standards to network and automation systemshas demonstrated that in practice, the parallel use of the frameworks and standards of the IEC 62357 series causes various problems – mainly due to their heterogeneity. Within the scope of its activities, the IEC TC 57 must combin aspects of IT automation and automation systems with each other. This involves a clash of technology paradigms. Information technology with techniques such as object-oriented modelling, service-oriented architecture (SOA), enterprise message bus systems (EMB) and semantic web technologies such as RDF (Resource Description Framework) and OWL (Web Ontology Language) on the one hand must be reconciled with conventional control and safety systems on the other; the latter focus on hard real-time requirements, communication and control-related security requirements, narrow bandwidths in communication and data point modelling.

One conflict results from the highly different perspectives of the technology domains in electricity supply under consideration here. Specifically, it must be noted that while the two main standard families of the IEC TC 57, Seamless Integration Architecture – the family of IEC 61970/61968 standards which documents the Common Information Model (CIM) and the IEC 61850 family for communication networks and systems in electrical substations – must be used seamlessly in the context of electricity supply, this proves impossible in reality.

Integration Of Standards

In this context, when further interfaces between the standards and frameworks at both the vertical and horizontal levels of integration are examined, further weaknesses related to the integration of standards become clear. In market communication, for example, (semantic) integration of XML messages (eXtensible Markup Language) in CIM format with the ebXML (Electronic Business XML) format is impossible as there are basic differences in modelling regarding concepts and syntax. Similar integration problems also exist in other domains, such as health care.

One possibility of solving the problems related to the overlapping of, and interfaces between, the levels of standards would be to agree on a common technology and semantic for harmonization. However, this is only possible in theory. On the one hand, the overall harmonization of such a system would be “virtually” impossible, i.e. involve unreasonably high costs and efforts, while on the other hand, the passed frameworks and standards have already been implemented in numerous products successfully placed on the market by their manufacturers. A basic change would be in violation of the security of investment for manufacturers and clients that has been made possible and promised by standardization. Given this, alternatives must be provided for this field.

While various types of heterogeneity occur in the IEC TC (Technical Committee) framework, not all of them can be eliminated. In the case of computers, for example, semiotic heterogeneity cannot be eliminated, whereas syntactic heterogeneity between formats can very well be remedied with the help of adapters and converters. Both terminological and conceptual heterogeneity are challenging. Given this, one sub-objective of this article is to prove by means of selected case studies that syntactical, terminological and conceptual heterogeneities can be reduced with the help of ontology mediation so that the frameworks can be integrated while existing standards can be maintained.

Interoperability – A definition

The IEEE defines interoperability as follows: “the ability of two or more systems or components to exchange information and to use the information that has been exchanged”. This article defines the issue of reducing syntactical, terminological and conceptual heterogeneity in order to improve the interoperability between systems that use the standards in energy industry. The issue of standard integration that is under consideration here must thus be regarded as a branch of interoperability research. This applies in particular to establishing semantic interoperability between components and systems. In research, the necessity of formal integration of the two standard families IEC 61850 and IEC 61970 has already been discussed. Integration of the two most important smart grid automation standards has already been achieved to some extent. However, testability and certification in particular require the seamless integration of these two standards in line with the IEC SIA, which must still be achieved.

The “COLIN” framework (Fig. 2) examines and solves the problem of various heterogeneities of the IEC TC 57 standards and related standards at technical level. In this context, various types of heterogeneities occur, including syntactical, terminological, conceptual and semiotic heterogeneity. The framework thus aims at demonstrating that these heterogeneities can be reduced by using ontologies as design artefacts, thus enabling mediation between the standards while maintaining the current versions.

Integration

The two families of standards CIM and 61850 play a particularly important role for standardization in the electricity sector. The development of these two standards was guided by different areas of focus. While the CIM family of standards focuses on control systems, distribution network management and business IT, the IEC 61850 standard addresses the automation and control of transformers, switchgears and distributed energy resources (DER). Both standards are extremely comprehensive and well tested. A problem is that both core families of standards from the IEC TC 57 were developed by two different groups with different focuses but must be used in the same context. Given this, integration of the standards must be implemented both in the form of static integration and integration at the time of configuration. This use case is designed to determine the extent to which established methods and tools are suitable to ensure these forms of integration and how an electronic model of the IEC 61850 in OWL format can contribute to integration.

The use case is aimed at creating an alignment between the data models of the IEC 61850 and the CIM. Yet so far there has been no reference alignment that could be used to evaluate the results. Given this, the use case needs to create not only an optimized matching (Fig. 3), but also a manual alignment to evaluate its own matcher. Established ontology matchers such as H-match, falcon AO or COMA++ may be used as an additional reference. In this context, an alignment must be prepared for both the general static data model and the SCL meta model, which is based on UML and can be serialized by means of XML. The aim is for the alignments and electronic models of the standards to permit use as design artefacts in various systems, and in particular within the scope of system integration by means of EAI.

Before starting preparation of the XAT tool, steps had to be taken to ensure that all standards were also available as electronic models in suitable serialisation. This proved to be the case for the CIM standard, which was available both as an ontology in OWL and RDF format and as a UML model in XMI format. The IEC 61850 family of standards was available as an XML-based model of LN and CDC developed by the IEC TC 57 WG 10 and was transformed into an ontology in a first step. Given this, both families of standards were then available as ontologies or XML models. In the next step, the models were used as the basis of a quantitative analysis to determine modelling focuses in the individual UML packages or LN groups. The more attributes and classes in a certain function, the more detailed its modelling and the more relevant it is compared to other parts of the standard. Applying this methodology, the use case succeeded in determining subsets of the standards that can be mapped to each other. As these subsets have identical higher-level functionalities, they also offer higher probabilities of a matching.

The tool’s two windows on the left show where the class is located in the model, the attributes it has, how many of these attributes are inherited and the number of children of the element. In addition, the tool also loads detailed descriptions from the standards. This information will then be used later on in matching. The tool also permits loading of various standards. For example, as shown on the right-hand side of the window logical nodes from the IEC 61850, here the nodes of the A group (automatic control function) can be loaded in addition to the CIM standard. Once the standards have been loaded, matching can be performed on the models. The tool also has a generic interface which permits loading of matchers created for the purpose in question. In addition, an element-based string matcher was implemented in the tool with the following strategy. Both the CIM and the IEC 61850 nodes have class and node names and the pertinent descriptions of class functions. While the CIM has a highly complex structure, the IEC 61850 is very simply structured in the form of a hierarchical taxonomy. Given this, structural information via ontologies was not really helpful and string-based matching was adopted as the goal.

Similarities Between Automation Standards

A stop word list was used to simplify the descriptions and perform additional stemming. The names of the CIM classes and of the logical nodes of the IEC 61850 standard were chosen to reflect their functions. Given this, the matcher uses weightings, with a matching of the names being weighted higher than a matching of the description or a matching of attributes. However, due to the naming conventions of the IEC 61850 standard, names are invariably short (four letters for the concept identifier, with the first letter representing the class), we aimed to select an algorithm able to identify duplicates in relatively short strings of characters. Based on the second-string library of the Carnegie Mellon University, we used a Java implementation of the Jaro-Winkler distance which supplies a normalised similarity between the strings (Fig. 4).

The three values for the individual string variables (classes, descriptions, attributes) are weighted differently and summarised to an overall similarity. After completion of the matcher, the tool provides feedback that the alignment was either saved in serialized form in the INRIA Alignment format or as OWL alignment as described in the introductory chapter. The alignment determined can be used in further tools, for example as input for a matcher or SCL tool.

Outlook

Using the COLIN Framework, we succeeded in developing a methodology for integrating the IEC standards CIM and IEC 61850 which maps the individual data models to each other on the basis of the existing electronic models. This methodology enables bijective mapping of the data of both standards. The mapping was documented as an electronic, explicitly formal mapping.

Based on this mapping, the two most important automation standards in the smart grid can be integrated in accordance with NIST, JISC, IEC or German DKE. This enables end-to-end communication within the IEC SIA TR 62357. Owing to differences in aggregation and granularity, bijective mapping is not possible for all test cases. Given this, there is still a high demand for the testability and certification of the individual standards to offer a quality of the original data for the mapping that is as high as possible.

The OPC Unified Architecture OPC UA is an automation technology of the future which will be able to use the prepared mapping within an OPC UA address space in the future. Based on the CIM or IEC 61850 standard, an address space mapping will be saved in the OPC UA server and used in enquiries to translate. This enables the server to ‘speak’ seamlessly to both CIM and IEC 61850.

TUV provides testing and expertise for Smart Energy through its Embedded Systems and Testing Laboratory (see sidebar, “TUV SUD Embedded Systems and Testing Laboratory for Smart Energy Technology”). TUV has established an framework to evaluate future energy grids (see sidebar “TUV SUD’s Expertise For The Energy Grids Of Tommorow”).

Sidebar: TÜV SÜD Embedded Systems and Testing
Laboratory for Smart Energy Technology

ÜV SÜD’s Smart Grid Centerof Excellence supports companies throughout the world, providing end-to-end strategic consulting services from planning to implementation. Technical testing of components and systems carried out in TÜV SÜD’s in-house testing laboratory for smart energy technology makes up the second service segment. The laboratory offers an environment to carry out simulation testing for both industrial and private applications and is equipped for the testing of both embedded systems and communication networks. The service portfolio covers testing of devices and systems, comprehensive consulting services on the design and development of new devices, smart-grid integration of components and systems and smart-grid optimization as well as offering specialized training for electricity producers and consumers.

• Assessment of available technologies for conformance with IEC 61850

• Development of an overall concept

• Component interoperability assessment

• Smart-grid integration

• Individual cost-benefit analyses

• Identification of grid infrastructure data

• Monitoring of compliance with IEC 61850 requirements

• Support during the implementation of smart grid-enabled systems

• Ongoing support and professional training of energy users and producers