ccss-sandia/rpowerflow

Steady-state electric power flow program written in Ruby

rpowerflow's Issues

14 bus test case does not converge

On my system, the bundled 14 bus test case does not seem to converge (the 6 bus case is fine).

If I force it to stop after 20 iterations, the results seem to be plausible:

#<struct Struct::Bus id=1, voltage=1.06000000055872, angle=-4.71494016630225e-09, mw=0.0, mvar=0.0, slack=true, avr=true, susceptance=0.0>
#<struct Struct::Bus id=2, voltage=1.04499999956269, angle=-0.184161940530395, mw=0.183, mvar=-0.127, slack=false, avr=true, susceptance=0.0>
#<struct Struct::Bus id=3, voltage=1.00999999969742, angle=-0.383413699708764, mw=-0.942, mvar=-0.19, slack=false, avr=true, susceptance=0.0>
#<struct Struct::Bus id=4, voltage=0.992435253636291, angle=-0.394286717422354, mw=-0.478, mvar=0.039, slack=false, avr=false, susceptance=0.0>
#<struct Struct::Bus id=5, voltage=0.992927841529518, angle=-0.374416124217786, mw=-0.076, mvar=-0.016, slack=false, avr=false, susceptance=0.0>
#<struct Struct::Bus id=6, voltage=1.06999999904211, angle=-0.864895155376082, mw=-0.112, mvar=-0.075, slack=false, avr=true, susceptance=0.0>
#<struct Struct::Bus id=7, voltage=1.00812572299551, angle=-0.536320001677928, mw=0.0, mvar=0.0, slack=false, avr=false, susceptance=0.0>
#<struct Struct::Bus id=8, voltage=1.08999999941587, angle=-0.536320001677928, mw=0.0, mvar=0.0, slack=false, avr=true, susceptance=0.0>
#<struct Struct::Bus id=9, voltage=0.971797467458968, angle=-0.612332692886332, mw=-0.295, mvar=-0.166, slack=false, avr=false, susceptance=0.0>
#<struct Struct::Bus id=10, voltage=0.975764364423406, angle=-0.664240538394374, mw=-0.09, mvar=-0.058, slack=false, avr=false, susceptance=0.0>
#<struct Struct::Bus id=11, voltage=1.01428071958757, angle=-0.767765110887086, mw=-0.035, mvar=-0.018, slack=false, avr=false, susceptance=0.0>
#<struct Struct::Bus id=12, voltage=1.15390792596165, angle=-1.62113363956345, mw=-0.061, mvar=-0.016, slack=false, avr=false, susceptance=0.0>
#<struct Struct::Bus id=13, voltage=0.840617836145766, angle=-0.746507939658906, mw=-0.135, mvar=-0.058, slack=false, avr=false, susceptance=0.0>
#<struct Struct::Bus id=14, voltage=0.891029874242489, angle=-0.70057286789995, mw=-0.149, mvar=-0.05, slack=false, avr=false, susceptance=0.0>

But if I let it go to 30 iterations things start to look pretty crazy:

#<struct Struct::Bus id=1, voltage=1.06003516311719, angle=1.88173589311591e-05, mw=0.0, mvar=0.0, slack=true, avr=true, susceptance=0.0>
#<struct Struct::Bus id=2, voltage=1.0449408611271, angle=17418306.5729259, mw=0.183, mvar=-0.127, slack=false, avr=true, susceptance=0.0>
#<struct Struct::Bus id=3, voltage=1.00996579021627, angle=10789483.3560132, mw=-0.942, mvar=-0.19, slack=false, avr=true, susceptance=0.0>
#<struct Struct::Bus id=4, voltage=-15713963.2312974, angle=49950564.0665059, mw=-0.478, mvar=0.039, slack=false, avr=false, susceptance=0.0>
#<struct Struct::Bus id=5, voltage=-17871317.3300304, angle=56215062.3127227, mw=-0.076, mvar=-0.016, slack=false, avr=false, susceptance=0.0>
#<struct Struct::Bus id=6, voltage=1.0701378047143, angle=-73223427.6012067, mw=-0.112, mvar=-0.075, slack=false, avr=true, susceptance=0.0>
#<struct Struct::Bus id=7, voltage=41160196.3234325, angle=7734746.19389931, mw=0.0, mvar=0.0, slack=false, avr=false, susceptance=0.0>
#<struct Struct::Bus id=8, voltage=1.08998966085348, angle=7734746.1938986, mw=0.0, mvar=0.0, slack=false, avr=true, susceptance=0.0>
#<struct Struct::Bus id=9, voltage=15647920.6518688, angle=-28586421.4830803, mw=-0.295, mvar=-0.166, slack=false, avr=false, susceptance=0.0>
#<struct Struct::Bus id=10, voltage=15416363.2353067, angle=-37926132.0788068, mw=-0.09, mvar=-0.058, slack=false, avr=false, susceptance=0.0>
#<struct Struct::Bus id=11, voltage=16901697.1985411, angle=-37392794.1367721, mw=-0.035, mvar=-0.018, slack=false, avr=false, susceptance=0.0>
#<struct Struct::Bus id=12, voltage=-2388632.54698258, angle=-9824987.57135228, mw=-0.061, mvar=-0.016, slack=false, avr=false, susceptance=0.0>
#<struct Struct::Bus id=13, voltage=14746581.707827, angle=-41680014.5759168, mw=-0.135, mvar=-0.058, slack=false, avr=false, susceptance=0.0>
#<struct Struct::Bus id=14, voltage=9328481.54405143, angle=-30633105.7909815, mw=-0.149, mvar=-0.05, slack=false, avr=false, susceptance=0.0>

If I let it go to the default 100 iterations, results vary. Sometimes I wind up with output where all of the above bus values are listed as "NaN", other times it crashes out with:

/opt/local/lib/ruby/gems/1.8/gems/rpowerflow-0.5.1/lib/rpowerflow/solvers/full_newton.rb:119:in `lu': singular matrix, status=1 (ZeroDivisionError)
from /opt/local/lib/ruby/gems/1.8/gems/rpowerflow-0.5.1/lib/rpowerflow/solvers/full_newton.rb:119:in `solve'
from /opt/local/lib/ruby/gems/1.8/gems/rpowerflow-0.5.1/lib/rpowerflow/solvers/full_newton.rb:47:in `times'
from /opt/local/lib/ruby/gems/1.8/gems/rpowerflow-0.5.1/lib/rpowerflow/solvers/full_newton.rb:47:in `solve'
from 14_bus_system_test_MAC.rb:103

I'm not sure what causes it to sometimes wind up one way and sometimes the other.

I'm on OS X 10.7.4, ruby 1.8.7 (2012-06-29 patchlevel 370), narray 0.6.0.1.