Application of PSO algorithms for VPP operation optimization

Abstract

The  random  character  of  generation  of  distributed energy  resources  and  their  increasing  numbers  cause  problems with  power  balancing  in  distribution  grids.  One  approach  to mitigate  the  balancing  issue  is  deployment  of  a  virtual  power plant   VPP   with   accordingly   dimensioned   components,   i.e. renewable generation and storages. It is also necessary to control the  power   flow  in  a  way  that  production  and  consumption balance  would  not  be  threaten.  From  an  investor's  perspective, profits  and  costs  are  important.  The  components  of  a  VPP installation,  such  as  generators  and  storage  systems  must  meet the  conditions  for  optimal  economic  and  technical  performance. Due to the increasing number of optimization variables, and thus the  complexity  of  the  objective  function,  it  is  necessary  to  use advanced  tools  for  effective  solution  for  optimization  problems. The  paper  presents  the  optimization  of  PV  and  electric  energy storage  for  selected  operational  scenario  using  particle  swarm algorithms.    The    proposed    method    allows    for    substantial improvement  of  power  losses  in  the  power  grid  and  decrease  of necessary   investment.   Predefined   power   consumption   target profiles  were  reached  even  with  reduced  PV  power  and  storage capacity.  Real  production  and  consumption  curves  were  used  in the presented analysis.