# ELECTENG 704 ASSIGNMENT #

FIND A SOLUTION AT Academic Writers Bay

ELECTENG 704 ASSIGNMENT #1 Date of Submission: 2020/9/22
1. In this assignment, we consider the rotary inverted pendulum shown in Figure 1 as the controlled
plant. The mechanism includes an arm attached to a motor, and a pendulum attached to the
arms tip. The rotation angles are detected by means of encoders installed at the base of the arm
and pendulum; these values are fed back to the arm position control in order to maintain the
pendulum in the inverted state. The rotary inverted pendulum is a 1-input 2-output system.
Figure 1: Rotary inverted pendulum
The Lagrangian equations of motion for the arm and pendulum are given as follows:
M(q)¨ q + Vm(q; q_) _ q + G(q) = Bu
where
B = 1 0 ; q = q q1 2 ; M(q) = Z1cos(Jq11 – q2) Z1cos(Jq21 – q2)
Page 2 of 4
ELECTENG 704 ASSIGNMENT #1 Date of Submission: 2020/9/22
Vm
(q; q_) = -µp – Zµ1psin +(µqa1 – q2) _ q1 -µp + Z1sin µp(q1 – q2) _ q2 ; G(q) = – -Z Z2 3sin sin( (q q1 2) )
with
J1 = Ja + mas2 a + mpla2; J2 = Jp + mps2 p; Z1 = mplasp; Z2 = fmasa + mplagg and Z3 = mpspg
The parameter values are,
g = 9:81 m=s2; ma = 0:275 kg; la = 0:322; sa = 0:175 m; Ja = 4:52 × 10-3 kg m2; mp = 0:065 kg;
Jp
= 1:36 × 10-3 kg m2; sp = 0:196 m; µa = 9:29 × 10-3Nms=rad; µp = 5:23 × 10-4Nms=rad:
(a) (15 points) Verify that the vertical position, q1 = 0, q2 = 0, _ q1 = 0 and _ q2 = 0, is an
equilibrium. Linearise the system about this equilibrium. Verify that the equilibrium is
unstable.
(b) (15 points) Suppose all the state variables can be measured. Design a LQR controller to
stabilise the system about the equilibrium. Check your result in a nonlinear simulation
with several initial conditions slightly off the vertical equilibrium. For this stabilising
controller, estimate the domain of attraction. HINT: Use different initial conditions.
(c) (15 points) Now assume that only q1 and q2 can be measured. Design a LQG controller to
stabilise the system about the equilibrium. Estimate the domain of attraction and compare
that with your LQR design.
(d) (15 points) Same assumption as (c) design a LTR controller to stabilise the system about
the equilibrium. Estimate the domain of attraction and compare that with your LQG
design.
(e) (15 points) Same assumption as (c) design an H1 controller to stabilise the system about
the equilibrium. Estimate the domain of attraction and compare that with your LQG
design.
2. Consider a system represented by the following linear continuous-time state equations:
x_ = Ax + Bu
where x 2 Rn is the state vector, u 2 Rm is the input vector, A and B are are known matrices
of appropriate dimensions. The objective is to determine the control input u which minimises
the following performance index (LQR Problem):
J = Z01 xTQx + uTRu dt;
where Q 2 Rn×n is real symmetric positive semi-definite matrix and R 2 Rp×p is a real positivedefinite matrix. The optimal control input which minimises J is given by,
u(t) = Kx(t); K = R-1BTP
where the matrix P is obtained by solving the is obtained by solving the following Riccati
equation.
ATP + PA + PBR-1BTP + Q < 0; P > 0; R > 0
Page 3 of 4
ELECTENG 704 ASSIGNMENT #1 Date of Submission: 2020/9/22
(a) (15 points) Using Schur compliment show that above inequalities can be represented using
following linear matrix inequalities. HINT: The Riccati equation, in contrast to Lyapunov
equations, is a nonlinear equation of P. This is because the quadratic term PBR-1BTP
appears in the inequality. This can be simplified using Schur compliment.
P > 0; Q > 0; R > 0 and
ATP +BTPA P + Q PB -R < 0:
(b) (10 points) For following A and B matrices find the optimal control gain K. HINT: Use
MATLAB LMI Solver or YALMIP toolbox with any solver.
A = -12 1 -1 ; B = 1 0
Page 4 of 4

YOU MAY ALSO READ ...  Dairy industry