WIND TURBINES MAINTENANCE MANAGEMENT BASED ON FTA AND BDD

Abstract

The  condition  of  a  wind  turbine  will  depend on different  elements  that  compound  it.  The  relationships  cause-effects  of  the  different  faults  can  be  analysed  qualitatively  by Fault  Tree  Analysis  (FTA).  FTA  is  a  graphical  representation of logical relationships between events, where each event has a fault probability associated. In this paper is proposed the quantitatively study of the FTA via Binary Decision Diagram (BDD). BDD is a method  which  determines  the  output  value  of  the  function  by examining  the  inputs.  The  BDD  method  does  not  analyse  the FTA directly, but converts the tree to the Boolean equations that will   provide   the   fault   probability   of   the   top   event.   This conversion   presents   several   problems,   where   the   variable ordering  scheme  chosen  for  the  construction  of  the  BDD  has  a crucial  effect  on  its  resulting  size.  It  is  solved  in  this  paper employing  the  Level,  Top-down-Left-Right,  AND,  Depth  First Search  and  Breadth-First  Search  methods  for  ranking  the  events (or vertices), and a comparative analysis is done.