To access the translated content: 1. The translated content of this course is available in regional languages. For details please visit https://nptel.ac.in/translation The video course content can be accessed in the form of regional language text transcripts, books which can be accessed under downloads of each course, subtitles in the video and Video Text Track below the video. Your feedback is highly appreciated. Kindly fill this form https://forms.gle/XFZhSnHsCLML2LXA6 2. Regional language subtitles available for this course To watch the subtitles in regional languages: 1. Click on the lecture under Course Details. 2. Play the video. 3. Now click on the Settings icon and a list of features will display 4. From that select the option Subtitles/CC. 5. Now select the Language from the available languages to read the subtitle in the regional language.
Namaskar, so in last class we were discussing about the state variable and we also saw that
if you have n order differential equation, so you can convert that nth order differential
equation into n first order equations and then we also discussed that these differential
equations which are first order in nature can be considered as a state equation and
we can define the various state variables and accordingly we can sum up the state variable
into a equation called a state equation and we also stabilize that the equation you can
write in the matrix form as x dot is equal to Ax plus Bu. We also discuss that the output
equation we can write in the similar fashion and we can write y is equal to Cx plus Du.
So this is what we discussed in the last session. Today we will discuss about the state diagram
and before going to the state diagram let us first understand the transfer function
in case of multivariable system that means if you have multiple variable in the system
how you will find out the transfer function. Now the definition of transfer function is
easily extended to a linear system with multiple input and outputs and this type of system
where you have multiple inputs and outputs we call them as multivariable systems.
Now if a linear system has p inputs and q outputs the transfer function between jth
input and the ith output can be defined as Gij in S domain equal to Yis divided by Rjs.
Now this is applicable when the other input variables are 0 that means Rks equal to 0
for any value of k not equal to j. So we are considering one input at a time.
Now this is only for considering the jth input in finding out the transfer function. If you
consider all the inputs say there are p inputs if you consider all the inputs you can write
the output transformed output in S domain as Gi1s R1s plus Gi2s R2s and so on up to
G1ps Rps. Now similarly for various outputs that is i ranging from 1 to q you can write
these equations and when you compile you can write in matrix form as Ys is equal to Gs
Rs. What is Ys? Ys is the number of outputs so it will range from 1 to q so that is why
the Ys will be q is to 1 output vector. Similarly Rs is the p is to 1 input vector.
Then what will be the value of G? G will be q into p matrix having the individual transfer
function corresponding to in relating the input and output of the number of components
we have in the inputs and outputs vector. So this is how you define the matrix which
contains the multiple output and multiple input. Now let us see when you have the multiple
input output that means we have a multivariable system how we will find out the transfer function.
So if you see this block diagram here you have one multivariable system where are multiple
inputs starting from R1 to Rp and outputs are from 1 to q. So in short you can say that
it has a vector R as an input and vector Y as an output. So now if you are asked to find
out the transfer function which is having a feedback loop how we will define in case
of multivariable input. So this Rs is multivariable input so you will have R as a vector as an
input and Y as an output containing the vector of q terms.
Now Gs is the transfer function so you can see that this can be considered as a transfer
function matrix which we just saw in this case. So what we can write so in this case
this is your the forward path gain matrix this is your feedback path gain matrix. Now
if you find out the value of Ys you can simply write that is Gs into Us, Us you can write
as Rs minus Bs and Bs you can write as Hs, Ys. So as we discussed Ys is q is to 1 output
vector and Us, Rs and Bs are p is to 1 vector then Gs and Hs are q into p or and p into
q transfer function matrices.
So dimension is given in case of Gs and Hs for Gs it is q into p and Hs is p into q.
Now the objective is that we need to eliminate the intermediate variables that is U and B.
So when we eliminate U and B from these 3 set of equations we can write Ys is equal
to Gs, Rs minus Gs, Hs, Ys. Now when you recompile Ys you can write as I plus Gs, Hs,
I is nothing but identity matrix. So this will become I plus Gs, Hs whole inverse into
Gs, Rs. Now the transfer function is nothing but Ys upon Rs. So if you say Ys upon Rs is
nothing but Ms. So you will get the closed loop transfer function matrix as 1 plus Gs,
Hs whole inverse into Gs. So this is how you will get the transfer function matrix in case
of multivariable system.
Now let us take one example suppose forward path transfer function matrix is Gs given
and the feedback path transfer function matrix is Hs it is given. Now we need to find the
closed loop transfer function matrix. So first we will try to find out the value of I plus
Gs, Hs. So if you put the value of I, I is nothing but it is identity matrix. So you
have this information in your hand. Gs, Hs is given. So when you put the values and simplify
you will get I plus Gs, Hs as s plus 2 upon s plus 1 then minus 1 by s 2 and s plus 3
upon s plus 2. Now next is we need to find out the transfer function matrix. So in that
case when we are finding the closed loop transfer function matrix first we have to take inverse.
So when we take inverse of the I plus Gs, Hs we can simply write 1 upon delta into s
plus 3 upon s plus 2 1 by s minus 2 s plus 2 upon s plus 1 and then we put the value
of Gs. Now the value of delta we can get from the matrix and when we simplify we get the
transfer function matrix as s into s plus 1 divided by s square plus 5 s plus 2 and
in bracket we will have one term as 3 s square plus 9 s plus 4 divided by s into s plus 1
into s plus 2 minus 1 by s then 2 and 3 s plus 2 by s into s plus 1. So this is how
you will get the transfer function matrix in case of multivariable system.
Now let us talk about the state diagram. So what is a state diagram? A state diagram is
nothing but the extension of the signal flow graph which we discussed in previous lectures
and this shows the state equations so which we discussed and the differential equation
because when you talk about the first order differential equation that is nothing but
the state equation we are talking about. So state equation and differential equation
can be shown in the extension of signal flow graph that is nothing but the state diagram.
Now the significance of the state diagram is that it forms a close relationship among
the state equations, the computer simulation and the transfer function. How? We will see
when we discuss the state diagram in detail in this session.
The state diagram is constructed following all rules of signal flow graph which we have
seen till now using Laplace transformed state equation. Now the difference between signal
flow graph and the state diagram is the integration operation which we have little bit difference
as compared to signal flow graph and the state diagram. So in this case when we talk about
the state diagram let us first see how the integration operation you will define in
the state diagram. Let us see there are two variables x1 and x2 and these are related
with each other as dx1 by dt is equal to x2. So if you represent them in the state diagram
you will say that this is basically x2 this is x1. So you can write dx1 by dt is nothing
but s into x1. So s into x1 will be giving you the value of x2. Now this is for this
when you are talking about the integration what we have to do we have to first take the
integration at on both sides we can write x1 t is nothing but equal to the integration
will range between t0 to t and x2 tau d tau plus x1 t0. So this is basically the initial
condition which we have taken at time t is equal to 0 time t is equal to t0.
Now if you take the Laplace transform so as we know that signal flow graph algebra does
not handle integration in time domain. So what we have to do we have to take the Laplace
transform on both sides. So we can write x1 s is nothing but Laplace of the integration
from t0 to t x2 d tau and plus x1 t0 by s. Now if you are doing the integration of this
you can separate this into two parts. So you can simply write one integration will range
between 0 to t x2 d tau and another is from 0 to t0 that is minus and integration between
0 to t0 x2 d tau. So this can be equivalently represented with the help of the integration
which we have broken into two parts and then x1 t0 by s which is the initial condition.
Now if you see the Laplace transform of this you can simply write it as x2 by s by the
definition of Laplace transform which we discussed previously. Now next item which is left is
Laplace of 0 to t0 x2 tau d tau plus x1 t0 by s. Now since the past history of the integrator
is represented by x1 t0 and the state transition is assumed to start at tau is equal to t0
means we are transiting at tau is equal to t0 that means x2 tau will be 0 before t0.
So what we can write x2 tau is 0 for the period between 0 to t0 it means that this particular
section you can replace by 0. So finally what you get x1 s is equal to x2 s by s plus x1 t0 by s
which is valid for time t greater than t0 because the period between 0 to t0 the value of the
initial condition for variable x2 is 0. Now the equation that is we have just found that is x1 s
is equal to x2 s by s plus x1 t0 by s is now algebraic and can be represented by the signal
flow graph. How will represent if you see this figure you have two variables one is x2 s another
is x1 s and we are summing up here that is summing block. So at this point this will be x2 s plus x1
t0 and then it is multiplied by 1 by s. So finally you get x1 s as x2 s plus x1 t0 divided by s which
is nothing but the equation which we just derived in case of the integration which we just performed.
Alternatively you can also represent it in a less number of connections. So say if this block is
representing x1 s so 1 by s you can put in both of the cases. So here when you sum up you will
get x2 s by s plus t1 plus x1 t0 by s. So both figures are representing the same equation. So
the advantage of this particular figure is that you need fewer element to show the equation which
we just derived. Now let us see how you will convert the differential equations into state
diagrams. So when the linear system is described by higher order differential equations the state
diagram can be constructed from these equations. How we will construct let us consider one
differential equation which is of nth order differential equation. So that is dny by dtn
plus an dn minus 1 by dtn minus 1 plus and so on up to a1 y equal to r. So this is suppose you have
got this differential equation and you have to convert this differential equation into state
diagram. So first what we have to do we have to construct the we have to rearrange this equation
as shown in this equation that is we will write dny by dtn equal to minus of an dn minus 1 y by
dtn minus 1 and so on minus a1 y t plus rt. So what we have done we have kept this as the
component and rest we have shifted to right side. So we got this equation from the previous equation.
Now what we can see we can see that the if you take the individual Laplace transform you can
write this as ys this particular component you can write as sys and this you can write as s to
the power n minus 1 ys and so on. So you know that individual components you can represent as
their Laplace transform. So what we will do we will this particular component you can replace
with the Laplace transform of rt. So we have now around total n element for one as the out
as the forcing function on the right side. So these are generally considered as a node.
So when you represent them as node how you will show them in the figure. So we will see why so
we have put one node as y then next is s into ys so we have put sy and so on. So we have got up to
sn minus 1 y and then sny for the last component and then we add r. So we have put all the elements
which we have seen in the in the equation as nodes. So what is sy ys so the ith order of the
differential component that is diy by dti is nothing but sy y. So for nth order component
we have written sny. Now we have defined the nodes next we use this equation to define the
links. So if you see the equation which we have just rearranged so what we can now see we can
see the connections. So if you see this particular node that is s to the power n y which is nothing
but representing this particular component that is dny by dtn. So that means this will be the
summation of all other components. So rt does not have any component coefficient there. So you will
connect r with s to the power n y with only unit gain. Now let us see this particular component.
So in s domain you can write as a minus a1 ys. So what will happen your coefficient for this
particular component is minus a1. So what you will do you will connect the y with s to the
power n y with minus a1. Similarly for this you can write minus a2 into s ys. So s ys you see
here and this is connected with s to the power n y as with the coefficient minus a2. So you got
minus a2 as the connection and similarly you connect all of them. So what you get finally
that is s to the power n y is nothing but r minus a n s minus s to the power n minus 1 y
minus a n minus 1 s to the power minus n minus 2 y and so on up to minus a2 s y minus a1 y.
So this particular figure represents the differential equation which we just written.
Now the important thing which still is left is that how we will correlate the nodes which are
connect which are adjacent to each other. How we will connect them? So just previous case when we
discuss we discussed that how x1 and x2 are related. So this knowledge we just discussed
will utilize in connecting the components. So if you see the component at this side x2 is
connected with x1 how you are connecting? You are connecting with s y is connected with y
with s to the power n minus 1 plus the initial condition for y that is y t0 by s. So this is
what you got from the equation which we just derived. So because if you see x1 x1 is nothing
but x2 by s into x1 t0 by s in our case x1 s is y x2 s is s y. So what we will do? We will use that
information and connect the adjacent nodes. So in this way you can connect all the adjacent
nodes with 1 by s and put the initial condition because when you divide the nth order differential
equation into n first order equations you will have the equation as we discussed in this case.
So you will get these equations in Laplace domain where you have initial conditions for
each and every state variable. So we will utilize that information and we will put the initial
conditions of all the state variable and connect them with 1 by s. So this is the final state
diagram which you will get and this will represent this particular differential equation.
Now let us take one example so that you can better understand. Let us see there is one
differential equation that is d2y by dt2 plus 3 dy by dt plus 2y equal to r. Now what we will do?
We will equate the highest order term of the last equation to the rest of the terms. So what we can
write? We can write d2y by dt2 is equal to minus 3 dy by dt minus 2y t plus rt. So here we will
have 1, 2, 3, 4 terms. So we will create 1, 2, 3, 4 nodes. Now this node is basically being
extended to give you better clarity. So you can add y with unit gain so that you can have the
clear understanding from the state diagram. So we have added these 4 nodes. One is r,
second is yt so you will replace with y, third is dy by dt so you will replace with sy and d2y by dt2
you will replace with s square y. Now what procedure we followed previously? We will
just use that procedure. So the y is adding to s square y with minus 2 as a coefficient so we have
minus 2 as a coefficient. dy by dt you multiply with minus 3 to add to s square y. So we get
these 2 connections and r is having 1 as a coefficient so r is connected to s square y
with 1 as a the gain. So now what we will do? We will connect these 2 nodes with each other
with the help of the integration function which we discussed. So we will connect with 1 by s,
s square y will connect it to sy with 1 by s and plus the initial condition you have for y1 t0 by
s and then for y we have y t0 by s as a initial condition. So finally you get the state diagram
for the differential equation which we just shown in the slide. Now the next question comes when
you are asked to find the transfer function from the state diagram. So what you will do?
So the transfer function between input and output is obtained from the state diagram by
using the gain formula and setting all other inputs and initial state to 0. So what we will
do? We will set the initial states to 0 and then we will equate input to output. So rs and ys is
obtained by using the gain formula. So what we will do? We can write s square y equal to minus
3 sy so that is from this minus 2y plus r. So what you can do? You can club all y component
one side when you club you will get s square plus 3 sy plus 2 into y equal to r. So this
is what you get when you check the signals which are coming to s square y. So when you sum up you
get this. Now what is the transfer function? Transfer function is y by r. So what you get
here? You will get that is s square plus 3s plus 2. So basically the y we have already taken out
so this will become s square plus 3s plus 2 into y equal to r. So y by r you get as 1 upon s square
plus 3s plus 2 which is nothing but the transfer function you have got from the state diagram.
So with this we can close our today's session in which we discuss mainly the state diagram. Now
we will see next how we will use the state diagram to find out the state equations. Thank you.