Page 147 - Decision Making Applications in Modern Power Systems
P. 147

112  Decision Making Applications in Modern Power Systems




                                          i NL1      i NL2
                                  i DGs          Z L1            Z Lg
                               Z i1
                                                                   i Grid
                 L f 1   L f 2    L f n       Load 1    Load 2
              i DG 1
                      i DG 2   i DG n
                 R f 1   R f 2    R f n
                                            NLL 1     NLL n
                                              1 ∂  n ∂  V PCC  i PCC
               S3 1  S4 1  S3 2  S4 2  S3 n  S4 n
                                         i DG1               I nll-local 1  I nll-local n
                                         i DG2      λ 1
                                              MOMPC   P and Q
               S  1 1  S2 1  S  2 1  S2 2  S  n 1  S2 n  λ n  calculation
              V DC 1   V DC 2  V DC n                        Low bandwidth
                                         i DGn
                                                  i fund
                                                i h
                DG 1  Z e1  DG 2  Z e2  DG n  Z en          communication
                                      Switching signals  Main controller
            FIGURE 4.11 A prototype smart grid with computational intelligence and communication
            links.
            where f 1 ½k 1 1Š, f 2 ½k 1 1Š; ... and f n ½k 1 1Š are related to different control
            objectives that should be minimized in the cost function in accordance with
            their weighting values ðλ 1 ; λ 2 ; ...; and λ n Þ in the cost function. It should
            be mentioned that the application of mentioned weighting factors will define
            the priority of each control objective. MOMPC could be applied to several
            MFDGs that operate in parallel; this is the main advantage of the MOMPC
            to handle several objectives at a time for several converters. These objectives
            could be reference tracking, fundamental and harmonic power sharing, power
            management, output current THD minimization, switching frequency control,
            etc. Since the idea behind MOMPC is to fulfill multiple objectives in an
            acceptable way and not having the best performance over an objective and
            affecting the other objectives in an inverse way, to fulfill the control objec-
            tives in an acceptable way, a decision-making should be done over defining
            weighting factors. This decision-making could be based on heuristic methods
            or learning, which is dependent on system smartness. An example of
            MOMPC could be applied to the microgrid shown in Fig. 4.11; the control
            objectives would be tracking the defined current reference to compensate the
            nonlinear load harmonics, fundamental and harmonic power sharing, and
            switching frequency control.
               The multiobjective cost function for this purpose will include three con-
            trol terms: the first term for reference current tracking, the second one for
            fundamental and harmonic power sharing, and the third term will control the
            switching frequency. The cost function will be as follows:

                  g½K 1 1Š 5 λ 1 3 I DG1 1 I DG2 2 I ref 1 λ 2 3 @ 1 I DG1 2 I ref



                                                                      ð4:10Þ
                           1 λ 3 3 f sw ðS sw ðkÞ; S sw ðk 1 1ÞÞ
            where λ 1 , λ 2 , and λ 3 are the weighting factors defining the priority of each
            control objective. And @ 1 is the power-sharing factor that in this case defines
            the power-sharing ratio between first and second MFDG and is defined as
   142   143   144   145   146   147   148   149   150   151   152