yanex.co.il
Products & solutions
Vision
Company



Grape - a software package for parallel and interactive
scheduling of equipment units to maintenance.

Operation Reliability and Economic impact

Grape is the sole software package for overall planning .
that is, maintenance schedule along extended periods, say a year.
Grape encompasses preventive maintenance, since forecast
failure data can be used as input constraints in Grape's optimization.

Grape deals with high-level decisions.
It ensures long-term operation Reliability,
having a great economic impact.

Interactive Working Cycle

Graphical entities are directly manipulated in the display.
Units are shifted along the time axis, locally recalculating optimal values.
One can fix only selected units, relaxing the remaining ones
to undergo another cycle of the parallel optimization.

The horizontal axis displayed is the planning period in time units,
say weeks. The vertical axis is the production quantity.
Equipment units are shown in different colors.
Each unit's left edge is the starting date and right edge
the end-date for maintenance outage.

The accumulated area of unit outages is the decrease in available production capacity.
Demand and threshold graphs are also shown for analysis.
Forecast and actual production can be shown and compared with proposed schedules.

Maintenance Scheduling Optimization

One is given a number of power plants, their units,
economic indices (fuel prices, maintenance costs, etc.)
and maintenance constraints (maintenance duration,
order of maintenance for sets of units, holidays excluded, etc.).

The desired output is a list of maintenance outages for
the planning period [say one year], subject to optimization criteria.
with reserve units to take care of forced outages.

Maintenance scheduling is known as a hard computational problem.
Grape's solution obtains the best in very short runs.

Production Tested in the Electric Power Industry

Grape solves a major management problem of power
plants and companies, as well as on a national level.
This is critical in an age of deregulation and fierce competition
and in geographical areas of fast growth of demand for electric power.

Specific optimization criteria for the electric power industry are
EUE [Expected Unserved Energy] and LOLP [Loss of Load Probability] thresholds.
Accordingly, the vertical axis of the above graph is power in MegaWatts.
In the above graph, peaks are during spring and autumn seasons, as expected.

The Israel Electric Corporation (IEC) - reports annual savings of $10 million
from improved maintenance scheduling using Grape.

GRAPE - a Computing Space compatible product

The parallel optimization has integer programming as an essential component.
Grape uses an algorithm to distribute tasks among processes,
such that they mutually accelerate each other, while working.
One gets results in minutes instead of hours.

Grape runs upon the platforms in the market.
Platform independence is guaranteed as it was developed using PCSpace.
Therefore it is Computing Space compatible.
 
© YANEX Computing Spaces 2004
HOME | Legal | Contacts | Human Resources