Problem formulation and analysis for optimal energy management in multisources systems: W effect

• Ghaith Warkozek ^[1] ; Stéphane Ploix ^[1] ; Frédéric Wurtz ^[1] ; Mireille Jacomino ^[1] ; Benoit Delinchant ^[1]
  1. [1] University of Grenoble
• Localización: Compel: International journal for computation and mathematics in electrical and electronic engineering, ISSN 0332-1649, Vol. 31, Nº 3 (Special Issue: Problems in Electromagnetism), 2012, págs. 904-919
• Idioma: inglés
• Enlaces
  • Texto completo
• Resumen

  • Purpose – The purpose of this paper is to introduce a problematic phenomenon that can occur when managing multi electrical sources systems by optimization.

    Design/methodology/approach – The energy management problem is formulated as a linear optimisation problem. Two approaches are developed and applied to detect the possible existence of equivalents solutions. The first is based on Dulmage‐Mendelsohn (DM) decomposition. With this method the structure of the optimisation problem is analysed. The second approach is a numeric approach; the detection of equivalents solutions is made by the formulation of new optimisation problem and the objective function of this problem is to maximise the distance between two equivalents solutions.

    Findings – The numeric approach is more efficient than the structural approach. In some cases, applying DM decomposition may not be sufficient to detect the risk of W effect. This is because DM decomposition does not take the value of variable's coefficient into consideration, which is important to determine the degrees of freedom in the set of variables.

    Originality/value – Multi sources systems are widely used, especially in buildings where renewable energies have good potential application. The linear formulation of the management problem may induce an existence of equivalent command strategies. The detection approach presented in this paper shows that some solutions are better than others from an applicabability point of view. They will not exhaust rapidly the storage system. This approach can be implemented in virtual sources plant to avoid solutions with this so‐called W effect.