• DocumentCode
    3063832
  • Title

    A solution to the unit commitment problem via molecular mechanics

  • Author

    Roa-Sepulveda, C. ; Pavez-Lazo, Boris ; Canales, Mauricio ; Lillo-Saavedra, Mario

  • Author_Institution
    Dept. de Ingenieria Electr., Univ. de Concepcion, Chile
  • Volume
    3
  • fYear
    2003
  • fDate
    23-26 June 2003
  • Abstract
    This work proposes a 4-state generating unit definition for the unit commitment (UC) formulation plus the standard definitions. This consideration produces a dimensionality problem that is tackled through the imitation of the geometrical optimisation problem of molecular structures to reach its minimum energy state. Based on classical mechanics laws, the geometric optimisation (GO) process is imitated, where molecular structures can be optimised from an energy viewpoint. To achieve this, the electric power system (EPS) is modelled as a molecular system (MS) composed of several virtual molecules where each one of them represents a generating unit. Thus, getting the minimum potential energy (via the van der Waals energy function) of the molecular system is equivalent to obtaining the minimum production cost of electric energy. To solve the problem, the simulated annealing (SA) technique is used as the optimisation technique. The EPS-MS model and method proposed here are validated in this work when calculating a 24-hour UC and its results are compared to an optimal solution analytically obtained and a solution using the conventional on/off states.
  • Keywords
    power generation economics; simulated annealing; van der Waals forces; electric power system; four-state generating unit definition; geometric optimisation process; geometrical optimisation problem; minimum energy state; minimum potential energy; minimum production cost; molecular mechanics; optimisation technique; simulated annealing technique; unit commitment problem; van der Waals energy function; virtual molecules; Cost function; Energy states; Irrigation; Lagrangian functions; Potential energy; Power generation; Power system modeling; Production systems; Simulated annealing; Solid modeling;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Power Tech Conference Proceedings, 2003 IEEE Bologna
  • Print_ISBN
    0-7803-7967-5
  • Type

    conf

  • DOI
    10.1109/PTC.2003.1304508
  • Filename
    1304508