Archiving software spots trends on the trains

Edited by the Engineeringtalk editorial team Sep 11, 2006

An out-of-the-box data acquisition and archiving package is ensuring that Austria's electric trains are kept at peak efficiency, while reducing the engineering time involved in their maintenance.

Austrian railway group OBB is responsible for managing the country's 10,000 km of rails, plus its 750 electric locomotives and 250 electric railcars.

With a typical locomotive weighing 86 tonnes, having a maximum speed of 230km/h and a traction of 270kN, each of its four traction motors requires periodic maintenance and testing to ensure peak efficiency.

In 2004 the company decided to call in outside expertise to carry out a major overhaul project for all of its motors.

Following an extensive period of evaluation, OBB awarded the contract to Lenze Anlagenbau.

Originally a component distributor for traction technology suppliers, in recent years Lenze has specialised in supplying full traction systems and is now the market leader in Austria.

Because of the scale of the project, Lenze's first job was to install six different load test beds at its test facility.

Four are for checking the traction motors, one carries out load tests and the sixth is a sinus test bed energised by an external dynamic convertor.

An added challenge is the fact that there is no industry standard for the motors: many different models with different performance characteristics are involved.

In order to carry out the project as efficiently as possible and to save engineering time, a visualisation and logging solution was required that is simple to use, includes metrological recording of electric and physical values, archives performance data, and that graphically represents performance data while logging test runs.

Lenze selected zenOn 6.01 from Salzburg-based Copa-Data to achieve this.

An extremely flexible, out-of-the-box solution, the zenOn 6.01 archive server records process data permanently and then archives it.

An archive can contain as many type-independent variables as required.

A particular advantage of zenOn in this project is that it provides three different types of archiving.

Firstly there is spontaneous, where on each change of a variable, the changed value is recorded.

Unnecessary entries caused by fluttering values are avoided through a hysteresis effect.

The second type is cyclic.

Here the values of an archive can be recorded down to a minimum cycle time of one second.

The third advantage is event triggering.

At the positive edge of a trigger bit, all values of an archive are recorded.

zenOn divides archives cyclically to individual archive files - this way an archive does not grow endlessly large and following archives stay computable.

The time interval is freely configurable.

Also, through functions archives can be started and stopped, for example with a shift or batch change.

In order to save storage space, archives can be exported automatically in data formats such as XML, CSV, or dBase, as well as copied onto backup systems or simply deleted.

But zenOn does a lot more, too.

Take, for example, following archives.

The sum, average, maximum and minimum are calculated from the archives over a freely selectable timeframe, and the resulting value is saved in a new archive.

The concise data preparation of the values - data reduction, automatic calculation of the average from the data, such as a recorded hour - makes it easier to detect trends.

The zenOn archive server also has perfect command of multiproject administration.

In an archive of an overriding project, values from a subproject are archived.

This way values from multiple facility sections and various projects can be stored together and optimally compared.

The supplying zenOn server can even be a CE-terminal.

All values within a heterogeneous system (PC and CE) can be seamlessly recorded in a central archive.

Then there is batch archiving, which allows for the easy allocation of batch designations to an archive.

The batch designation is recorded in a string variable and saved with the archive.

In the zenOn modules 'Extended Trend' or 'Report Generator' the designation can conveniently be filtered.

So far the project has been running successfully for two years, to the satisfaction of both OBB and Lenze.

The concise data preparation of the values makes it easier to detect trends, and the motor-specific evaluation of historic data ensures that individual peak efficiencies are maintained.

The result is that Lenze is hoping to add extra test beds in the near future.