Electrical Network 3-5 Graph Theory Properties of tree 1) Tree contains all nodes on the graph. 2) Tree does not contain any closed path. 3) In a tree, there exists only one path between any pair of nodes. 4) In a tree, minimum end nodes or terminal nodes are two. 5) Every connected graph has atleast one tree. 6) The rank of the tree is same as the rank of graph ic. (n ~ 1) 7) Tree contain (n - 1) branches if n are the nodes of the tree. 3.1.7 Branch of a Tree (Twig) A branch of a tree is called twig. If there are n number of nodes on a graph then the tree of a graph contains (n - 1) twigs. Then referring Fig. 3.7 (b), twigs of a tree can be written as follows. For first tree, twigs are T = {(2), (5), (6)} For second tree, twigs are T = {(2), (4), (5)} For third tree, twigs are T = {(1), (4), (6)} 3.1.8 Cotree A set of branches forming a complement of tree is called cotree. The number of branches of a cotree equals b - (n — 1) where b is number of branches of a graph. 3.1.9 Chord (or Link) The branches which are not in a tree are called chords or links. Refer Fig. 3.7 (b) For first tree, chords or links are L = {(1), (3), ()} For second tree, chords or links are L = {(1), (3), (6)} For third tree, chords or links are L = {(2), 3), 6)} 3.1.10 Loop In general, a network consists of ‘n’ nodes which are interconnected in some way by 'b’ branches, then it is possible to traverse adjacent branches starting at any node and return to the original starting node in different ways. Such a closed path formed by the network branches is called as loop. Loop is a connected subgraph of a connected graph such that at each node there are two branches incident. If the two terminals are made to coincide, it will generate a loop or circuit. The Fig. 3.8 shows different possible loops.Electrical Network 3-6 Graph Theory B @) ¢ (3) (5) 6 “@ 8 4 Cc D (a) (b) Fig. 3.8 Loops Properties of loops A loop of a graph has following properties. 1) There are exactly two paths between any -pair of nodes in the circuit, 2) There exists atleast two branches in loop. 3) The maximum possible branches in a loop are equal to number of nodes. 3.2 Incidence Matrix An oriented graph can be completely explained or represented with the help of a matrix known as incidence matrix. The incidence matrix gives information about the branches incident on which nodes and their orientations at the node (whether towards or away from node). There are two types of incidence matrix namely, (@) Complete incidence matrix (A,) (ii) Reduced incidence matrix (A). 3.2.1 Complete Incidence Matrix (A,) The incidence matrix is nothing but a mathematical model to represent the given network with all the information available. The information regarding the network is nothing but which branches are incident at which node and what are the orientations relative to the nodes. All this information is written in a matrix form which is called complete incidence matrix. It is represented by A,. Consider following network and its oriented graph as shown in Fig. 39. (b) Fig. 3.9 Network and oriented graphElectrical Network 3-7 Graph Theory There are 4 nodes A, B, C, D and 6 branches as 1, 2, 3, 4, 5 and 6. The directions are selected randomly for explanation purpose. For a graph , if there are ‘n’ nodes and 'b’ branches, the complete incidence matrix A, = ajy is a rectangular matrix of order _n xb. The standard conventions to be followed while writting the incidence matrix are as given below. |, if branch k is associated with node h and oriented away from node h. ~ 1, if branch k is associated with node h and oriented towards node h. 0, if branch k is not associated with node h. ang The complete incidence matrix of the network shown in Fig. 3.9 (b) is as given below in the tabular form, Nodes = A,B,C,D Branches = 1, 2, 3,4, 5,6 Nodes 4 2 3 4 5 6 A a 1 1 ° 0 ° 8 o a 0 “1 4 oO c 0 0 a 4 0 a D 1 0 ° 0 a = The above result can be written in matrix form as follows, 1 2 3 4 5 6 ¢€ Branches -1 +1 +1 0 0 07A L Nodes Aa =/0 -1 0 -1 + O|B 0 0 -1 +41 0 +H14C +10 0 0 +4 -1)D In above matrix, the-summation of elements in any column results in zero value. This is because each of the branch is considered in one column. Each branch connects two nodes. The current flows towards the node at one end of the branch, while it flows away from the node at other end. Any one row of the complete incidence matrix can be obtained by algebraic manipulation of other rows. For example, in above matrix, the fourth row is negative of sum of first three rows. Atleast we can eliminate one of the rows as that can be obtained by negative of sum of others. Properties of complete incidence matrix Each row in the matrix corresponds to a node of the graph. Each row has non-zero entries such as +1 and -1 depending upon the orientation of branch at the nodes. Also the entries in all other columns of that row are zero.Electrical Network 3-9 Graph Theory A 8 c For a given graph select a tree. In the incidence matrix, arrange the columns such that the first (n- 1) @ “ columns correspond to the twigs for the selected tree and (5) last (b — n + 1) columns correspond to the links. Suppose we select one tree as shown in Fig. 3.11. 5 Then reduced incidence matrix A can be written by writing columns corresponding to twigs first and columns Fig. 3.11 Tree corresponding to chords (links) after that. Twigs Links . 245 136 aAz=[1 9 0 2 1 0] 1-411 000) 010 0-3 1} The reduced incidence matrix can be written as, A= [Ar AL] Where, Ay is the square matrix of the order (n ~ 1) x (n - 1) whose columns corresponds to twigs, hence name of the matrix is twig matrix denoted by Ay. A, is the matrix of the order (n - 1) x (b - n + 1) whose columns represents the links, hence name of the matrix is link matrix denoted by A,. Number of possible trees of a graph For a given graph, it is possible to calculate number of possible trees of the network using reduced incidence matrix. The number of possible trees for any linear graph is given by expression given as follows. Number of possible trees of a linear graph = det ([A] [A]"), Where, [A] = Reduced incidence matrix {A]T_ = Transpose of a reduced incidence matrix. mp Example 3.2: For the given oriented graph write complete incidence matrix. Also write reduced incidence matrix. Determine number of possible trees for given oriented graph. ae B 4 ogElectrical Network 3-10 Graph Theory Solution : In the given oriented graph, Nodes = A,B,C, D Branches = 1,2,3,4,5 Hence the complete incidence matrix is given by, 1 2 3 4 5 ©-Branches -1 1 0 0 07A 4 Nodes Aa =} 0-1 1 1 O/B 0 0 O-1 1/C€ 1 0-1 0 -14)D In above matrix, 4** row is negative of sum of first three rows. Hence eliminating 4" row, the reduced incidence matrix is given by, 1 23 45 © Branches aufi 10 0 07a ¢ Nodes 0-11 10\B 0 00-11]JC The number of possible trees is given by, Number of possible trees = det {[A] [A]"} The product of reduced incidence matrix [A] and its transpose matrix [A]" is given by, -1 0 0 -1 10 00 1-1 0 2-1 0 [AMAT =| 0 -2 1001 Of] 0 1 Of =; 3 A 0 00 -21J/{0 1 -1 o-1 2 oo. Hence number of possible trees is given by, 2-10 (6- 1)+ 1(-2)+0 Number of possible trees = det (A][A]"} =| -1 3. -1 0-2 0-1 2| =8Electrical Network 3-11 Graph Theory 3.3 Loop Matrix (or Circuit Matrix) 3.3.1 Loop Matrix (B,) For a given directed graph, it is always possible to state which branches are involved in the formation of loops or closed circuits or meshes. For a given oriented graph, loops can be identified by marking each loop with orientation of loop current by a curved arrow. This enables to write the nodes in the loop in cyclic order. In complicated networks with many loops, it is advisable to list the nodes in selected cyclic order to avoid the confusion between oriented loop currents. A loop matrix or circuit matrix is represented by B,. For a graph with n nodes and b branches, loop matrix B, is a rectangular matrix with b columns (equal to number of branches) and as many as rows as there are loops. The elements bj, of loop matrix has following values. by = + 1, if branch k is in loop h and its orientation coincides orientation of loop. = - 1, if branch k is in loop h and its orientation does not coincide. = 0, if branch k is not in loop h. Consider an oriented graph as shown in the Fig. 3.13. Fig. 3.13 Orionted graph with loop orientations For above oriented graph various possible loops can be listed as follows. Circuit 1 : (1, 2, 3] Circuit 2 : [3, 4, 5] Circuit 3 : [2, 6] Circuit 4 : (1, 2, 4, 5] Circuit 5 : [1, 6, 3] etc.Electrical Network 3-12 Graph Theory Above listing can be expressed in tabular form as follows. Circuits Branches 1 2 3 4 5 6 1 1 1 1 ° ° ° 2 0 0 -1 1 1 0 3 ° -1 0 0 ° 1 4 1 1 oO 1 1 0 1 oO 0 This circuit matrix B, is given by, 1 2 3 4 5 6 & Branches 1 1 1.0 0 0741 J Circuits or Loops p,- [0 0-211 0/2 0-1 0001]3 11 011 044 10 100145 ‘The rank of the circuit matrix B, is given by [b ~ (n ~ 1)}. Thus for a network with many loops, the size of the circuit matrix increases. To ease simplification, a smaller subset of the circuit matrix is considered which is known as fundamental circuit (f-circuit) matrix. 3.3.2 Fundamental Circuit or f-Circuit or Tieset Matrix (B) It is a subset of circuit matrix. It is represented by B. For a given graph, select any tree and remove its all links. Then replace each link one at a time, it will form a closed path or loop or closed circuit along with some of the tree branches existing. The circuits formed by replacing each link in tree are called fundamental circuits or f-circuits or tiesets. It is very important to select the orientation of E-circuit to coincide with that of the link completing the f-circuit. The number of f-circuits equals the number of links. For a graph with 'b’ branches and 'n' nodes, the possible f-circuits are given by [b — (n - 1)]. Consider a tree of oriented graph as shown in the Fig. 3.14 (b). ‘As shown above, twigs 2, 3, 4 are forming tree while 1, 5, 6 are the links which are not the part of tree. When these links are replaced, we will have three f-circuits as shown in the Fig. 3.14 (0), (d) and (¢). The orientation of each loop is coinciding the orientation of the connecting link in each f-circuit. The f-circuits or tiesets are given by writing link entry first and then other branches in sequence as follows,Graph Theory é A ‘Bog c A 2 Bo4 Cc 3 ’ ‘s D D @ ® fs avis 4c 2 3 D © C) © Fig. 3.14 f-circuit 1: [1, 2, 3] or f-tieset 1: [1, 2, 3] or floop 1: [1, 2, 3] f-tieset 2 : [5, 3, 4] or f-loop 2 : [5, 3, 4] ftieset 3 : [6, 2] or loop 3: (6, 2) We can write this information about f-circuits in tabular form which is known as tieset Schedule. The tieset schedule for above f-circuits is as Feircuit 2: [5, 3,4] or f-cireuit 3: [6, 2] or feircuits Branches 1 2 3 4 6 1 1 1 1 ° ° ° 2 oO oO -1 1 1 0 3 0 -1 Oo oO oO 4 Hence f-circuit matrix B is given by, 1 2 3 4.5 6 © Branches p-f1 1 1.0 0 0)71 J f-circuits 0 0 111 04/2 0-1 000143 In general, matrix B is arranged such that first columns corresponding to entries of links and then columns corresponding to entries of links and then columns corresponding to entries of twigs as given below.Electrical Network 3-44 Graph Theory 156 2 34 100110 B=/0100-411 001-100 Links Twigs = (By: By] = U:By Where, B,, is the matrix corresponding to link entries called link matrix. It is always identity matrix. B, is called twig matrix corresponding to twig entries. 3.4 Cutset Matrix and Fundamental Cutset Matrix A connected graph can be separated into two parts by removing certain branches of the graph. This is equivalent to cutting a graph into two parts hence it is referred as cutset. A cutset is a minimal set of branches of a connected graph such that after removal of these branches graph gets separated into two distinct parts, each of which a connected graph, with the condition that replacing any one branch from the cutset makes the graph connected. Consider an oriented graph as shown in the Fig. 3.15 (a). A 8 c 2 1 5 D (e) 6 A228 c 8 coA c z 4 2° 8 3 5 5 5 D D (a) @) co) Fig. 3.15Electrical Network 3-15 Graph Theory Consider that branches 1, 3, 4 and 6 are removed from graph which are shown by dashed lines in the Fig. 3.15 (b). The graph is divided into two parts each of which is connected. Now each branch is replaced one at a time, we get connected graph for every branch replaced as shown in the Fig. 3.15 (c), (d), (e) and (f). Thus a set of branches [1,3,4, 5] forms a cutset. We can have another set of branches [6, 2, 3, 5] giving another cutset. Hence for any graph, there may be more than one cutsets and every graph has atleast one cutset. The cutset separates the nodes of the graph into 2 group each being in one of the two parts. So the branch of cutset has one of its terminals incident at a node in one group and its other end at a node in the other group. So we can consider that all the branches joined at a node forms a cutset since by removing them, the graph would be splitted into two parts. 3.4.1 Cutset Matrix (Q,) The cutset matrix Q, can be written by considering the orientation of the cutset from one of the two parts to the other. Consider an oriented graph as shown in the Fig. 3.16 (a). Consider orientations of cutsets arbitrarily. The possible 6 cutsets are as shown in the Fig. 3.16 (b). 3 A 2 8 4 Ad ce suf, ct } 5 6, D o (a) Oriented graph (6) Oriantations of cutsets Fig. 3.16 The elements of cutset matrix Q, are given as follows, qy = + 1, if branch j is in the cutset i with orientation of both coinciding. = — 1, if branch j is in the cutset i with orientation not coinciding. = 0, if branch j is not in the cutset i. For the graph shown in the Fig. 3.16 (b), the cutsets can be listed as follows, Cutset - C; : [1,2] Cutset - C, : [2, 3, 4] Cutset - Cy : [4, 5]Electrical Network 3-16 Graph Theory Cutset - C, : [1, 3, 5] Cutset - Cs : [1, 3, 4] - Cutset - C, : [2, 3, 5] The cutset matrix can be written in tabular form, Cutsets Branches 1 2 3 4 5 6 cy 1 -1 0 ° 0 0 Cc, 0 1 -1 -1 0 0 C3 ° ° 0 1 -1 ° Cy -1 0 1 0 1 0 Cs 1 0 -1 “1 0 ° Cy 0 1 -1 0 -1 0 ‘Thus, cutset matrix Q, is given by, 1 2 3 4 5 © Branches 1-1 0 0 07C J Catsets 0123-21 0/G Q=]0 0 0 1 -A]e; 10310 1/cG 1 0 -1 -1 O}G 0 1 + 0 ASG For any graph, number of cutsets may be extensively large so to ease simplification subset of cutset matrix is considered which is called fundamental cutset or f-cutset matrix. 3.4.2 Fundamental Cutset (or f-Cutset Matrix) (Q) A fundamental cutset of a graph is a set of one twig and chords at a node. It is apparent that at each node for each twig, there will be a fundamental cutset. For a graph with n nodes, there will be n - 1 number of fundamental cutsets. The orientation of f-cutsets is selected such that it coincides with the orientation of the defining twig of f-cutset. Consider an oriented graph as shown in the Fig, 3.17 (a). Select one tree as shown in the Fig. 3.17 (b). The twigs are shown by dark lines while links are shown by dashed lines. The fundamental cutsets are given by (n - 1) ie. (4 - 1) = 3. The fundamental cutsets are as shown in the Fig. 3.17 (), (d), and (e).Electrical Network ~ 3-17 BD (a) Oriented graph foufset 3 0 (e) Fundamental cutsets Fig. 3.17 The f-cutsets can be listed as below. feutset 1 : [2, 1] feutset 2 : [4, 5] frcutset 3 : [3, 5, 1] We can write the information about fundamental cutsets in a tabular form which is known as cutset schedule. The f-cutset schedule is given as follows.Electrical Network 3-48 Graph Theory f-cutsets Branches 1 2 3 4 5 foutset 1 [2, 1) -1 1 0 0 o feutset 2 [4, 5] ° ° oO 4 a foutset 1 (3,5, 1] - 1 0 1 0 1 Hence fundamental cutset matrix is given by, 12 4 5 € Branches 3 Q- [72 1.00 0] f-cutset 14 f-cutsets 00.01 -1] f-éutset2 -1 010 1] f-cutset3 Many times fundamental cutset matrix is written by writing twig entries first and then link entries. The fundamental cutset matrix can then be written as, 243 1 5 q=fieoe.4 0 010 0-1 001+ 1 eS Twigs Unk [Qr :Qu] (U:Qu] Where Qr is the twig matrix with all entries of the twigs of the graph. Moreover twig matrix is identity matrix ie. 100 [Qr] = [U] = }0 1 0 ooi1 The matrix consisting entries of links is called link matrix denoted by Qu. 1m» Example 3.3: In the following network, the numerical values of resistances also indicate the branch numbers. Write the oriented graph of the network. Select a tree with branches 1, 2, 3 as the tree branches, write tieset and cutset schedule,Electrical Network 3-19 Graph Theory 6v 4) = 62 22» 30 Fig, 3.18 Solution : For the branch current direction shown, the oriented graph of a given network is as shown in Fig. 3.18 (a). ‘The tree with branches 1, 2, 3 is as shown in the Fig. 3.18 (b) 228 DY ‘ (a) Oriented graph (b) Tree of a graph Fig. 3.18 The loop currents I,, Ig, Ic are in the direction of the respective links of a tree. The tiesets are written as follows, Ftieset 1:1, [ 5, 1,2] ftieset 2: Ig [ 4, 3,1] ftieset 3: I, [ 6, 2,3] The tieset schedule can be written as follows, tiesots Branches 1| 2 3 4 8 6 | mB | -1 | -1 0 of+:] 0 | Iel.3] | +4 o | -1 | +4 o | o | Ic (6, 2, o} -1 +4 0 0 +1Electrical Network 3-20 Graph Theory Fig. 3.18 (c) The cutset schedule is written as follows : Now considering node ‘B’ as reference and also assuming orientation of cutset to be same as that of tree branches as shown in Fig. 3.18 (c). The cutsets are written as follows, f cutset 1 : A [2, 5, 6] f cutset 2 : B[3, 4, 6] f cutset 3: C [1, 5, 4] f-cutsets Branches 1 2 3 4 5 6 Al, 5, 6 0 ca 0 0 es = 8B. 4,6 0 0 4 Ht ° a Cis 5, 4 4“ 0 0 1 1 0 3.5 Orthogonal Relationship between Matrix A and B and Matrix Q and B For any linear graph, if we ‘arrange columns of the complete incidence matrix A, and loop matrix B, in the same order, then A, Br = B, Al =0 (1) Above equation (1) indicates that matrices A and 8, are orthogonal to each other. This property also: gets satisfied with the reduced incidence matrix A and fundamental loop or fundamental tieset matrix B. If the reduced incidence matrix A and tieset matrix B are obtained for the same tree, there exists orthogonal relationship between matrices given as, ABT = BAT = 0 w Q) Similary there exists orthogonal relationship between loop or circuit or tieset matrix B, and cutset matrix if the columns of both the matrices are arranged in same order. The orthogonal relationship between Q, and B, is given by, Qi Br = 2 Qa = 0 = @)Electrical Network 3-21 Graph Theory This property also gets satisfied with the fundamental tieset matrix B and fundamental cutset matrix Q. If both are written for the same tree. Then the orthogonal relationship between B and Q is given by, QBT = BOT = 0 ~ @) 3.6 Relation between Submatrices A, B and Q Consider that for a given tree, the reduced incidence matrix A, f-tieset matrix B and f-cutset matrix Q are arranged in such a way that all the twig entries are arranged first and then all the link entries. Then the reduced incidence matrix, ftieset matrix and f-cutset matrices can be expressed as follows interms of twig matrix and link matrix. The reduced incidence matrix = [A] = [Ar : At] ‘The f-tieset matrix = [B] = [Br : BL] = [Br : U] The f-cutset matrix = [Q] = [Qr :Q.] = [U:Qu] But we know that there exists orthogonal relationship betweeen A and B given as, Br ABT = [Ar:A = =0 wr aus ArBr +A BL = 0 ArBh = -A.BL = (Q) But A, ie. link matrix is non-singular ie. |A| #0 and thus Aj’ exists. Multiplying equation (1) Aj! BL = -Aj) ALBE Br = -By (Az'A,)T Hence matrix B can be written as, B= [-(Aj'-Ay)T:Ul w= 2) But we know that ABT = 0, ALBr = -Ar By Postmultiplying with (B])~ 1, we get Av = -Ar By Br)" Ay = -Ar Br @)T Av = ~Ar (By! By)Electrical Network 3-22 Graph Theory Hence matrix A can be written as, A = [Ar:-Ar @j!-Br)T] A= Aq (U:~(Bi!-Br)T) ~ @) Similarly, we can prove that, Q= [U:- Bc! -Br)T] ww (4) From equations (3) and (4) we can write, A= Ar-Q Q= AZA Q = AAT: ALI Q=[U:az! Au} ~ 8) But we have already proved that, By = -(Az!-AL)T BT = (Aq!-AUI » (6) Hence Q can be written as, Q = (U:-By Q =-B;! = 3.7 Network Analysis using Graph Theory To analyse a network means to find current through any branch or voltage across any branch. We can analyse given network using graph theo, by relating branch currents and loop currents or branch voltages and node voltages (across tree branches). 3.7.1 Relationship between Branch Currents and Loop Currents Let [I,] represents branch current matrix and [I,] represents loop current matrix, then we can relate branch currents and loop currents as, (i) = (87 th ~@ Where, [B"] is the transpose of fundamental circuit matrix [B].Electrical Network 3-23 Graph Theory When a current source is present across any branch then the total branch current will be the algebraic addition of branch current and branch input current of the source. Consider two conditions in a network as shown in the Fig. 3.19. Consider Fig. 3.19 (a). The direction of the branch current and current source is same. Hence effective branch current flowing from terminal A to B is given by subtracting source current as ly = -1 -Q) ANW- —w— ty ety on) 8 Abs é ©, —or— @ (») Fig 3.19 Consider Fig. 3.19 (b). The direction of the branch current and current source is opposite. Hence equivalent branch current flowing from terminal A to B is given by adding source current as, Ih = +h + GB) In such special cases, relationship between branch currents and loop currents can be modified interms of source currents as follows, 0) = (B01 + 0) ~- (4) Where, [I,] is a column matrix representing current source across each branch 3.7.2 Relationship between Branch Voltages and Node Voltages Let [V,] be the branch voltage matrix and [V,] be the node voltage matrix, then we can relate branch voltage and node voltage as 1V) = [Q11V,] + 6) Where, [Q"] is the transpose of fundamental cutset matrix (Q]. When a voltage source is present in series branch then the total branch voltage is the algebraic addition of source voltage and voltage across branch. Consider general conditions in any network as shown in the Fig. 3.20. @) ) Fig. 3.20Electrical Network 3-28 Graph Theory Consider Fig. 3.20 (a). If polarities of source voltage and branch voltage are same, then add source voltage and write branch voltage using sign conventions as, V, = Voltage across A - B = Vi + V, - 6) Consider Fig. 3.20 (b). If polarities of source voltage and branch voltage are different then subtract source voltage as, v= W-V, = @ Hence in such special cases we can modify the relationship between branch voltages and node voltages interms of source voltages as, Iva] = [QIV.] + IV + @) Where, [V,] represents a column matrix representing voltage sources in series with branches. 3.7.3 Network Equilibrium Equations on the Basis of Loop Analysis Inloop analysis, using graph theory, the minimum number of equilibrium equations required is equal to [b — (n — 1)] where 'b' is the number of branches and 'n' is the number of nodes in the graph. Consider a circuit as shown in the Fig. 3.21 (a) and an oriented graph of it as shown in the Fig. 3.21 (6). (b) Oriented graph Fig. 3.24 Let the branch currents in respective branches be ij, i, iy iy i, and ig with the orientations same as those of branches. Consider a tree for oriented graph as shown in the Fig. 3.22 (a).Electrical Network 3-25 Graph Theory (a) Fundamental circuits (b) Fundamental cutsets Fig. 3.22 Let I, I, and J, be the loop currents. The directions of loop currents are coinciding with the orientations of the links in the respective loops. We can list f-circuits as, f-cireuit 1: [1, 2, 5] ... (for loop current I,) f-cireuit 2 : [3, 4, 2] ... (for loop current I) fecircuit 3 : [6, 5, 4] ... (for loop current I;) Writing circuits in tabular form, Then fundamental circuit matrix B is given by, 1 2 3 4 5 6 ¢-Branches (Branch currents) py = [2 2 9 0 1 07 14 F-circuits 0-11-10 0/2 0 001-1143Electrical Network 3-26 Graph Theory The branch currents can be expressed interms of loop currents as, OJ = (7) 0 in 1 0 0 iz 1-1 0 is} _|0 1 0 r ig} [O =a 1 L is 1 0 -;'% ip oo 1 From above equation we can write, i, = Ive] -h- is =h, y= -h+lyis=h-ly k= The branch voltages can be expressed as product of branch current and branch impedance. Let vj, V2 ... Vg be branch voltages. "In the first branch we have voltage source V, in series with R,. As the polarity of voltage source and that of voltage across R, is not same hence subtracting source voltage y= he R-Vy In branch 2 there is no voltage source Vv) = ip-R, Again in branch 3, voltage source V; is present. The polarities of source voltage and voltage across R, are same hence adding source voltage. vy = is-Ry+ Vp In all remaining branches, there is no voltage source. Vy = ig :Ry Ys =ig/Rs and vg=ig-Ry Consider fundamental circuit matrix B again. Each «row in the matrix represents summation of branch voltages around fundamental circuits. Hence we can write, vitvatvs =0 -vatva-v, =O | By KVL va—Vo tv, =0 Putting values of V,, V2, V;, Vz, Vs and V, in above equations we can write, O ie. iy Ry + ig Ry + ig Ry = Vy ~ ig Ry + ig: Ry + Vp-ig-Ry=0 ie. —iy-Ry + ig-Ry—iy-Ry=- V2 w+ (10) i,-R,-is-R, +ig-R,=0 + (1) i, +R, - Vy + iy Ry + ig Rs,Electrical Network 3-27 Graph Theory Substituting branch currents expressed interms of the loop currents, T,-R, + 0, —1)-R, + 0, - 1) -Rs = Vy ,-R, + -R,-l-R,+1,-Rs-h-Rs = Vi (Ry +R, +R) I - R= Rs +13 = Vy vs (12) ~ G ~ T)- Ry + )- Rs - (- E) Ry = - Vn =I, -R,+1-R,+h-R,-b- Ry t+], -Ry-1,+(R, +R, +R)-L-R-l, = - Vp we (13) (ly - 1) -Ry- (ly - b) Rs + (I) Re = 0 Tye Ry- Bp Ry Tye Ry + 1° Ry + 1h“ Ry = 0 ~ Rs -Ry-L, + (Ry + Ry + Ry) 1, = 0 s (14) Equations (12), (13) and (14) represent equilibrium equations for the given network on the basis of loop analysis. R,=-V, 3.7.4 Network Equilibrium Equations on the Basis of Node Analysis Consider Fig. 3.23. The oriented graph is as shown in Fig, 3.23 (b). Assume the selected tree as shown in Fig. 3.24. (a) Network (b) Oriented araph Fig. 3.23 Network and oriented graph Node analysis is based on Kirchhoft’s current law. For the given tree, we can write cutset as follows. Feutset 2 : (2, 3, 1] Feutset 4 : [4, 3, 6] f-cutset 5 : [5, 1, 6] Fig, 3.24 Tree of a given networkElectrical Network 3-28 Graph Theory The above set of cutset can be written in tabular form as follows : —— Note that in above tabular form, the elements of a column represent branch voltages and elements of the row represent branch currents. Let branch voltages be denoted by E,, E,, Es, Ey, E; and Ey, Let the voltages at nodes A,C,D be Va, Vc and Vp where node B is selected as datum node or reference node or zero potential node. The branch voltages can be written as : Branch voltage E, (represented by column 1) = - V,- Vp Branch voltage E, (represented by column 2) = Vj Branch voltage E, (represented by column 3) = V,- Vc Branch voltage E, (represented by column 4) Ve Branch voltage E, (represented by column 5) = Vp Ve- Vp But for branch voltage E, there is voltage source V, present. Also there is voltage source V, present in branch 3. Branch voltage E, (represented by column 6) E, = V,-Va-Vp Also there is voltage source V, present in branch 3. s By = W+Va-Ve ‘As we have seen already, elements in a row represent a branch current and according to the Kirchhoff’s current law, the algebraic sum of the currents at any node is always zero, we can add all the elements of each row to get zero current at that node. At node A: -I,+h+1,=0 w= (15) At node C: -I, +1, +1,=0 o» (16) At node D: - I, +1;-k=0 = (17) = = =F But hag beg bogElectrical Network 3-29 Graph Theory Substituting all these values in above equations, we have E, Er, Ey -ByE, ete te To ie _ Ok-Wa=Mo) , May XE Roe hae +. (18) The second equation can be written as, oO Le. =0 we (19) The third equation can be written as, ~H, Es _ Es _ ie RRs "Re ec _Wx-Ma-Vo) , Yo , (e=Vd) © R +t R, 7° wes (20) The branch voltages E, E,, Ey, Ey, Es and E, can be expressed interms of node pair voltages V4, Ve and Vp. Thus equations (18), (19) and (20) will then become three independent equilibrium equations interms of variables V,, Vc and Vp. Then by solving all these equation simultaneously, we can get values of node voltages, branch voltages and branch currents. mm) Example 3.4: Write the tieset schedule for the network shown in Fig. 3.25 and using the tieset shedule obtain the equilibrium equations on loop current basis. 92 52 Fig. 3.25 Solution: In the above network, the branch current orientations are not given. Hence assuming randomly the branch current orientations as shown in Fig. 3.25 (a). We have to replace 5 V voltage source by a short circuit. The oriented graph of a given network is as shown in Fig, 3.25 (b).Electrical Network 3-30 Graph Theory A A 5Q, 5a by by 8) 5 c Fig, 3.25 (b) Oriented graph For writing tieset schedule, consider a tree of a oriented graph as shown in Fig. 3.25 (c). Let currents I, I, and I, be the loop currents in the orientation of b;, b3; and bg respectively. The fcircuits or loops X, Y, Z can be written as follows Fecircuit X ; [by, by, by] Fig. 3.25 (c) Tree of a graph f-circuit ry Ds, by} Feircuit Z : [by by bs] The tieset schedule can be written as follows, Loop currents Branches by | be | bs | be | bs | be X fb4, bo, bal ea] o | -1 ° 0 Y [bs, bs, bal 0 -1 | +1 oO 1 ° Z [bg by, bs] ° ° o fa | a fon The tieset matrix can be written as follows, Links Twigs +100 Ort 1 0 B=/0 #4 Of 0 # oot OH A> Let the currents I,, I, I, I, I, and I, be the respective branch currents. From tieset schedule, we can write 6 branch currents as follows, Lek h-L-lElectrical Network h=h key keh From the rows of tieset schedule, we can write three equilibrium equations as follows, V,+V,-Vy=5 -V,+V3+Vs =0 V,-Vs+V_ =0 Where V,, Vz, Vy, Vy Vs and Vz are the branch voltages respectively. Graph Theory .- (1) +» (2) + 3 Hence we can write branch voltages interms of the branch currents as follows, But But Substituting values of branch voltages V, = G) 6) hek Vv, =5i V2 = @) (10) h=@-l) V; = 10 (,-1,) V5 = (h) 6) -l V, = 51, V4 = ) 10) y= @-l) Vq = 100, - 1) Vs = (ls) (10) I, = d,-1) Vs = 10 (Il, - 1) Vv. = @) 6) k= Ve = 5h, in equations (1), (2) and (3). 51, +10 (,-1,)-10(,-1) =5 51, +101,-10,- 101 +101, =5 251,-101,- 101, =5 G@)Electrical Network 3-32 Graph Theory 100, ~ 1,) +5 1, + 101, 1.) = 0 “101+ 101,+51,+101,-101, =0 10 1, + 251,- 101, = 0 BAC) 0 0 100, -1,)-100,-1,) +51, 101,-101,- 101, +101, +51, = “101, - 101, +251, = 0 s+» Bla) Hence the equilibrium equations are as follows, 251,-101,- 101, = 5 101, + 251,- 101, = 0 -101,- 101, +251, =0 3.8 V-Shift and I-Shift Method A voltage source with a passive branch in series with it is called an accompanied voltage source. Similarly a current source with a passive branch zm ®: in parallel with it is called an accompanied current source. Such accompanied voltage and current Fig. 3.26 Network with accompanied sources S°urces are shown in the Fig. 3.26. While drawing a graph, such a accompanied source can be replaced by their internal impedances. If their internal impedances are not given, then replace voltage source by short circuit and current source by open circuit. But if in the network, the sources are unaccompanied, then it is very difficult to-select number of unknowns interms of number of branches if the sources are treated as separate branches. Consider a network shown in the Fig, 3.26 (a). % R, My |} ©) v R, L Ro L, 1 oR NOI vy TBO How. vl =o 1 (a) Network with unaccompanied sources (b) Network with voltage and current sources shifted Fig. 3.26Electrical Network 3-33 Graph Theory We can get some equivalence between the networks with and without accompanied sources by using source shifting technique in the network with unaccompanied sources. Thus by using source shifting technique unaccompanied source can be made accompanied source. Consider a network shown in the Fig. 3.27 (a). C, 2} Cc, AB A (a) Network with unaccompanied source (b) V-shift technique Fig. 327 In such a cases, short terminals A and B and push voltage source V through the node in series with passive element R, and R,, This makes unaccompanied, accompanied one. This is known as voltage source shifting technique or V-shift technique. The networks shown in the Fig. 3.27 (a) and (b) are equivalent as far as KVL is concerned. Consider a network shown in the Fig. 3.28 (a) in which current source ‘is unaccompanied. A A (a) Network with unaccompanied source (b) 1 - shift technique Fig. 3.28 In such a cases, remove original current source and introduce additional current sources which are identical to original one across each branch as shown in the Fig. 3.28 (b). The technique is known as current source shifting technique or I-shift technique. The networks shown in the Fig. 3.28 (a) and (b) are equivalent as far as KCL is concerned.Electrical Network 3-34 Graph Theory Thus V-shift and Ishift techniques are most useful in making all the sources ab accompanied sources and it is most suitable to consider all sources as accompanied for writing loop and node equations. 3.9 Network Equilibrium Equations in Matrix Form In last section we have studied how to obtain equilibrium equations for a network which are based on loop analysis and node analysis. We can represent the same equation in matrix form which simplifies the task of analysis of a large scale network. 3.9.1 Loop or Mesh or KVL Equilibrium Equations The KVL equilibrium equations based on loop or mesh analysis are given by, BZ, BT I, = B(V.-Zp I.) or Zu=ek Where, Z = B2,BT and E = B(V,~Zp Is) ‘The quantities involved in above equations are as follows. B is the f-circuit or tieset matrix of order (b - n+1) xb. Zp is the branch impedance matrix of order bxb. 1, is the column matrix of loop currents of order (b-n +1) x 1. V, is the column matrix of voltage sources of order bx1. I, is column matrix of source currents of order bx 1. Z is coefficient matrix also known as loop impedance matrix of order (b-n+1) x (b-n +1) E is the column matrix of order (bn +1)x1 which represents voltages in loop. The matrix equation ZI, =E represents a set of (b~n+1) independent loop equations for a network with n nodes and b branches. 3.9.2 Node or KCL Equilibrium Equation The KCL equilibrium equations based on nodal analysis are given by, Qs QTV: = Qs Vs) or YVYUst- where Y = Q¥%Q" and = Q(.-% V.) The quantities involved in above equations are as follows. Qis the fcutset matrix of order (n-1) xb.