Page 214 - Advances in Renewable Energies and Power Technologies
P. 214

3. Necessity of Joint Adoption of Distributed Maximum Power Point   187




                  to measure I SC to obtain the approximate IeV characteristic. The set of equations that
                  allow to obtain the approximate IeVand PeV characteristics of a string composed by
                  an arbitrary number N of buckeboost based LSCPVUs are given below:

                                    V cost $b$I SCk
                        if  I ¼ I 0k ¼
                                   V ds max   V cost
                                                                                (5.33)
                                                           P
                                            P
                                               V cost $b$I SCi  V ds max $b$I SCy
                                            i˛J            y˛T
                        V tot ¼ V ds max   V cost þ      þ
                                                 I 0k        I 0k þ b$I SCy
                  where k ¼ 1, 2, .,N;J ¼ {i > k: I 0i < I 0k }; T ¼ {y < k: I 0y > I 0k }.
                         if  I ¼ I costk ¼ I ds max   b$I SCk



                               2                                        3
                                                                                (5.34)
                                P              P
                                  V cost $b$I SCi  V cost $b$I SCi
                               6 i˛J           i˛J           V cost $b$I SCk  7
                         V tot ¼  6         ;              þ            7
                                    I costk        I costk      I costk
                               4                                        5
                  where J ¼ {k < i < N: I costi > I costk }, k ¼ 1, 2, .,N. In Eqs. (5.33) and (5.34), the
                  sets J and T may also be empty. Eqs. (5.33) and (5.34) represent the core of FEMPV
                  algorithm suited for buckeboost based LSCPVUs.
                  3.2.1.3 Buck-based Lossless Self-controlled Photovoltaic Units
                  Also, the IeVand PeV characteristics of a buck-based LSCPVU can be subdivided
                  into three portions. In particular, for 0   V   V 1 , the corresponding portions of the
                  IeV and PeV characteristics of the LSCPVU are straight lines of Eq. I ¼ I MAX
                  [54,55] for IeV curve and P ¼ V$I MAX for the PeV curve. Instead, for
                  V 1   V   V MPP , the corresponding portion of IeV characteristic of the LSCPVU
                  is a hyperbole of Eq. V$I ¼ P MPP [54,55]. This means that, for V 1   V   V MPP ,
                  the corresponding portion of PeV characteristic of the LSCPVU is flat and equal
                  to P ¼ P MPP . Finally, for V MPP   V   V oc the IeV and PeV characteristics of the
                  LSCPVU are coincident with those of the associated PV module [54,55]. The mean-
                  ing of V 1 and I MAX is the same as for the case of buckeboost based LSCPVUs.
                  Therefore, also in the case of buck-based LSCPVU it is:
                                                    P MPP
                                               V 1 ¼                            (5.35)
                                                    I MAX
                     Continuous Conduction Mode operating conditions of the buck converter and
                  related equations will be referred to in the sequel because of the synchronous version
                  of the buck which has been considered. Without the adoption of a proper overcurrent
                  protection circuitry, values of V lower than V 1 would give rise to output currents I
                  greater than I MAX . So that, for V   V 1 , the output current is limited to I MAX by
                  means of a proper overcurrent protection circuitry and the PV module operating
   209   210   211   212   213   214   215   216   217   218   219