Page 140 - Power Electronic Control in Electrical Systems
P. 140

//SYS21/F:/PEC/REVISES_10-11-01/075065126-CH004.3D ± 128 ± [106±152/47] 17.11.2001 9:54AM







               128 Power flows in compensation and control studies

                         y l and  jV l j are the incremental changes in nodal voltage magnitude and phase
                      angle at node l,
                        (r) represents the r-th iterative step and l ˆ 2, 3, 4, ... n.
                        The elements of the Jacobian matrix can be found by differentiating equations
                      (4.42) and (4.43) with respect to y l , y m , jV l j and jV m j.
                        For the case when l ˆ m:
                                n
                      @P l     X                                            2              2
                         ˆjV l j  jV m jf G lm sin (y l   y m ) ‡ B lm cos (y l   y m )g jV l j B ll ˆ Q l  jV l j B ll
                      @y l
                               m ˆ l
                                                                                        (4:45)
                              n
                       @P l  X                                               P l
                           ˆ    jV m jfG lm cos(y l   y m ) ‡ B lm sin(y l   y m )g‡jV l jG ll ˆ  ‡jV l jG ll (4:46)
                       @jV l j                                               jV l j
                             mˆl
                                 n
                      @Q l      X                                           2            2
                          ˆjV l j  jV m jfG lm cos (y l   y m ) ‡ B lm sin (y l   y m )g jV l j G ll ˆ P l  jV l j G ll
                       @y l
                                m ˆ l
                                                                                        (4:47)
                                n
                        @Q l   X                                                Q l
                             ˆ    jV m jfG lm sin (y l   y m )   B lm cos (y l   y m )g jV l jB ll ˆ   jV l jB ll
                        @jV l j                                                 jV l j
                               m ˆ l
                                                                                        (4:48)
                      For the case when l 6ˆ m
                                   @P l
                                       ˆjV l jjV m jfG lm sin (y l   y m )   B lm cos (y l   y m )g  (4:49)
                                   @y m
                             @P l                                           1 @Q l
                                 ˆjV l jfG lm cos (y l   y m ) ‡ B lm sin (y l   y m )gˆ   (4:50)
                            @jV m j                                        jV m j @y m
                                  @Q l
                                      ˆ jV l jjV m jfG lm cos (y l   y m ) ‡ B lm sin (y l   y m )g  (4:51)
                                  @y m
                                @Q l                                          1 @P l
                                     ˆjV l jfG lm sin (y l   y m )   B lm cos (y l   y m )gˆ  (4:52)
                               @jV m j                                      jV m j @y m
                      To start the iterative solution, initial estimates of the nodal voltage magnitudes and
                      phase angles at all the PQ nodes and voltage phase angles at all the PV nodes are
                      given to calculate the active and reactive power injections using equations (4.42±
                      4.43). Since it is unlikely that the initial estimated voltages will agree with the voltages
                      at the solution point, the calculated power injections will not agree with the known
                      specified powers.
                        The mismatch power vectors may be defined as

                                     DP (r)  ˆ (P gen    P load )   P calc,(r)  ˆ P net    P calc,(r)  (4:53)
                                    DQ (r)  ˆ (Q gen    Q load )   Q calc,(r)  ˆ Q net    Q calc,(r)  (4:54)
                      The Jacobian elements are then calculated and the linearized equation (4.44) is solved
                      to obtain the vectors of voltage updates
                                                 q (r ‡ 1)  ˆ q (r)  ‡ Dq (r)           (4:55)

                                                 (r ‡ 1)   (r)     (r)
                                               jVj    ˆjVj   ‡ DjVj                     (4:56)
   135   136   137   138   139   140   141   142   143   144   145