Wednesday, October 28, 2009

The "Semantic Web" in Power System Automation

Traditionally almost all communication solutions for automation and especially for power system automation is build on hundred and thousands of "Points" (Signals) organized in huge lists. Each "Point" has a type and a kind of a simple index (or identifier). Different vendors (or even different people) use different list. In one case the "Phase A Voltage to ground" may have the index "26717" in another case it may be "363.26". Do you know what these numbers mean?

This is comparable to the web of today: search engines are searching mainly for ASCII strings. If I search for "Guenther" "Wilhelm", I could not specify that "Guenther" is the first name and "Wilhelm" is the family name. Google returns 18,700,000 hits. Searching for "Guenther Wilhelm" returns just 30,000 hits. The second is closer to what I am looking for. I would like to search linke this: "firstname = Guenther" and "familyname = Wilhelm". In this example we have added some semantic (meaning) to the names.

It would be nice to have reasonable names for the "signals" instead of just numbers and to have semantic added to the "signals". This would allow to interpret the list of signals - IF THE NAMES ARE CHOOSEN TO MEAN SOMETHING USEFUL. IEC 61850 and IEC 61400-25 are standards that define semantic and names for each signal - like the Semantic Web does.

According to Wikipedia is "The Semantic Web an evolving development of the World Wide Web in which the meaning (semantics) of information and services on the web is defined, making it possible for the web to understand and satisfy the requests of people and machines to use the web content."

In IEC 61850 we have decided many years ago that the name for the three-phase electrical system should be the same all over (in principle) and the same for all voltage levels - because the electrical system is the smae all over (with different voltages and frequences). The following picture shows two voltage levels and a single model for the three-phase system. The name "MMXU" stands for a logical node defining all crucial information that describes a three-phase electrical system. The "PhV" (phase voltage) has a "PhsA" value etc. Each of the values has SI-Units, scaling factors etc. These names expose the same information allover, in all applications (in substations and in factories, on ships, on railways, ...). Why do we need myriads of different indices in current solutions for the same information?


The communication based on simple lists seems to be simple. But if your company has Millions of points to test and to manage ... what then? Guess there is no need to discuss the problems handling huge lists - lists that are differently formatted and contained in Wordfiles, Spreadsheets, pdf files, just on paper, data bases, ... How could one make these lists machine readable? One of my customers told me that the have to maintain 1,300 documents containing signal lists - wow.

IEC 61850 is - to my knowledge - the only comprehensive standard that defines common and specific information models for the electric power industry and beyond. We had a proposal to add a "FishCounter" for hydro power plants ... why not? The standard also defines services to exchange the values and concrete protocols to serialize the services.

IEC 61850 could be understood as the "Semantic Web" of the power automation and protection world. Now you can read the Phase voltage of MMXU1 of the logical device SpyDER under the address: In order to know where this device is located you just can talk to the device to retrieve some description or you can use the system configuration description file (according to IEC 61850-6). This file has all semantic information including the binding of the model to the real world.

Click HERE for an example of a device that implements the MMXU logical node and exposes the voltage of the power outlet it is connected to.

In this regard IEC 61850 is MAYA (Most Advanced Yet Accepted) -- accepted all over.

No comments: