TECNOLOGA Y AUTOMATIZACIN INDUSTRIAL (IN196)

EXAMEN FINAL
Ciclo 2016-01

Seccin : INA2
Profesores : George Power, Mirko Klusmann
Duracin : 110 minutos

Indicaciones:

No est permitido el uso de ningn material de consulta (cuadernos, libros, diccionarios,

lecturas, etc.).
No se permite el uso de calculadora cientfica o programable.
No se permite el uso de telfono celular o cualquier otro aparato electrnico durante el
examen, los cuales deben permanecer apagados y guardados.
Escriba con lapicero azul o negro solo en los espacios indicados
Los exmenes resueltos con lpiz no tienen derecho a reclamo
Puede usar las caras en blanco como borrador (no ser corregido)

1 (5 p.) 2 (5 p.) 3 (5 p.) 4 (5 p.) Nota

1
PART 1 (5 pts.)

Answer concisely the following questions regarding control algorithms:

a) Explain which control algorithms are considered continuous and discontinuous

Continuous = analog I/O signals, i.e. P, PI, PD, PID, also fuzzy and NN
Discontinuous = discrete I/O signals, i.e. ON/OFF, NN in some cases.

b) What are the disadvantages of proportional control?

P controller does not eliminate permanent error, need of readjustment with manual reset.

c) What is controller tuning?

The process of finding the gains for P, I and D actions according to system performance.
Different methods: trial and error, Ziegler-Nichols, auto tuning, etc.

d) In which situations is a PID controller not recommended?

Systems with slow response (e.g. temperature control in a furnace), unsteady input signals
(derivative action tends to oscillations) or in cases an on/off controller will suffice (no need
for stable output)

e) Predictive control is characterized by

The use of a mathematical model to predict the system output, because it is slow and/or
output signals cannot be measured, for example in a rotary furnace or a ball mill.

2
PART 2 (5 pts.)

An automotive speed controller designed with fuzzy logic uses 3 inputs (speed, acceleration
and distance to destination) and 1 output (power or fuel flow to engine), as shown on the fol-
lowing diagrams.

Figure 2.1 Fuzzification of input variables

Figure 2.2 Defuzzification of power output

Some example of inference rules are:

IF speed is TOO SLOW and acceleration is DECELERATING, THEN INCREASE
POWER GREATLY
IF speed is SLOW and acceleration is DECREASING, THEN INCREASE POWER
SLIGHTLY
IF distance is CLOSE, THEN DECREASE POWER SLIGHTLY

3
a) Explain the difference between crisp and fuzzy values
Crisp or sharp values have an exact numerical value obtained from the sensing system, i.e.
temperature = 245 degrees; fuzzy values are expressed in terms of linguistic variables, i.e.
temperature is warm which will be converted to a value in the range of 0 to 1 (degree of
membership) during fuzzification.

b) Establish three additional rules for this system

If distance is distant and speed is slow or too slow, then increase power greatly
If distance is close and speed is optimum, then leave power constant
If distance is very close and acceleration is decelerating, then decrease power greatly

c) Explain how the power output is calculated

Degrees of membership of each input variable are applied to all inference rules using
min/max transformations and calculating a value (0 to 1) for each output variable; all values
are combined during defuzzification and the centroid of all areas is the output value.

d) What are the advantages of fuzzy controllers compared to classic control algorithms?
They can be used to express uncertainty and the experience of operators, no need of com-
plex calculations and tuning.

e) Mention other applications of fuzzy controllers

A/C systems, robots, appliances (i.e. level and time control in washing machines), digital
photo cameras.

4
PART 3 (5 pts.)

A sequential control system to automate a drill is to be programmed. The setup of the drill is
shown and the process sequence are described below.

Figure 3.1 Setup of automated drill

The initial state of the drill is defined as follows:

The drill motor and cooling pump are stopped
The carriage/drill is in the upper position
There is no work piece in the clamping device

The entire drilling sequence can be divided into the following sections:
Insert work piece (manual)
If required, set switch for coolant (depending on the material)
Start the machine with the start button (drill motor starts up)
Clamp the work piece with the selected clamp pressure
Start the cooling pump (if coolant selected)
Lower drill and carriage to the bottom target position (drill)
Wait 0.5 seconds at lower target position (drill)
Raise drill with carriage to upper target position
Remove work piece, turn off drill motor and cooling pump
Remove work piece (manual)

5
 Figure 3.2 Time sequence of drill operation

Complete the table of inputs and outputs and design the program sequence in S7 Graph.

Inputs
Address Symbol Explanation
E 0.0 Dr_mot_running Feedback signal for drill running at selected speed
E 0.1 Dr_mot_stopped Feedback signal for drill stopped
E 0.2 Drill_down Limit switch for drill in lower position
E 0.3 Drill_up Limit switch for drill in upper position
E 0.4 Cl_press_ok Feedback signal for work piece clamping pressure reached
E 0.5 Coolant_sel Selector for coolant (dependent on work piece)
E 0.6 Cool_press_ok Feedback signal for coolant pressure reached
E 0.7 Start_button Start button of the drill

Outputs
Address Symbol Explanation
A 0.0 Dr_mot_on Turn on drill motor
A 0.1 Cool_pump_on Turn on cooling pump (dependent on work piece)
A 0.2 Lower_drill Lower drill and carriage to bottom limit position
A 0.3 Raise_drill Raise drill and carriage to the upper limit position
A 0.4 Clamp_workp Clamp work piece at required pressure

6
First level Graph:

7
Second level Graph:

8
PART 4 (5 pts.)
Figure 4.1 shows a robot arm configured for painting car bumpers. For each position on the
paint line the robot moves from the safe position and begins with the bumper on front and must
spray the paint along three lines and normal to the surface following the bumper shape (one line
above, one in the middle, one below). This is done three times, then the robot moves back to
the safe position and proceeds with the bumper on the rear. After finishing, the paint line moves
forward one position and the whole procedure is repeated.

Figure 4.1 Paint line with robot arm

a) Describe the points needed to program the robot for one paint position
At least 5 points are needed for each paint trajectory (see figure), 3 parallel trajectories are
needed to cover all the bumper (bottom, middle, top), one approximation position for
begin/end of each trajectory, one middle position and one safe (rest) position.

b) Which input and output signals are necessary?

Input 1: Bumper in position (transfer from paint line)
Output 1: Activate paint gun
Output 2: Paint position finished

9
c) Sketch the program instructions for painting one bumper

WAIT INPUT(1)=1 Wait for bumper in position

SPEED FAST
MOVE P(0, 1) Move close to left side of bumper
SPEED SLOW
MOVE P(1, 1) Move to first point of trajectory
SPEED MEDIUM
OUTPUT(1) = 1 Activate paint gun
MOVE P(2, 1) Paint first trajectory from left to right
MOVE P(3, 1)
MOVE P(4, 1)
MOVE P(5, 1)
OUTPUT(1) = 0 Turn off paint gun
SPEED SLOW
MOVE P(6, 1) Move close to last point of first trajectory
MOVE P(6, 2) Move close to last point of second trajectory
MOVE P(5, 2) Move to last point of second trajectory
SPEED MEDIUM
OUTPUT(1) = 1
MOVE P(4, 2)
MOVE P(3, 2)
MOVE P(2, 2)
MOVE P(1, 2)

etc.

Note: Better results with more points per trajectory or curve interpolation (depends on robot soft-
ware). Some robots can learn the trajectory from a sensor arm manipulated by an experienced
painter.

10