Case Study

Container Crane Control

<using FuzzyTech>

Objectives of Ports
For delivery of goods through containers transported by cargo ships. Example is PTP in Johore, Wesport in Klang and of course Singapore.

Crane Productivity
Crane productivity is measured by how fast the Port Authority can move the cranes. Singapore = 25 moves/hour Jakarta = 17 moves/hour Malaysia Westport=22 moves/hour

Crane Productivity in Westport, Port Klang

Container Crane Simulator

Container Crane Control

Loading and unloading of containers are done in harbors in every country around the world. For transportation of manufactured goods, food, etc. Container cranes are used for such purpose.

Operations and Problems

When a container is picked up and the crane head starts to move, the container begins to sway. Swaying of the container is not a problem during transportation but a swaying container cannot be released.

Container Crane Control

Two ways to solve this problem:
1. To position the crane head exactly over the target position, and then just wait until the sway dampens to an acceptable level. 2. To pick up the container and just move slowly that no sway ever occurs. Both ways would be alright on a non-windy day but it takes too much time. An alternative is to build container cranes where additional cables fix the position of the container during operation- but this would be too expensive.

 For these reasons, most container cranes use continuous speed control of the crane motor- a human operator then control the speed of the motor.

The operator has to simultaneously compensate for the sway and make sure the target position is reached in time. This is not an easy and would need very skilled operators.<You can try to manually control the crane through simulation>

Several Control Modes

Many engineers have tried to automate this control task of controlling the crane by using:
Conventional PID Control Model-based control Fuzzy logic control

Problems with PID

This is a nonlinear problem. Minimizing the swaying of the container is important when the container is closed to the target where PID is insufficient due to high nonlinearity.

Problems with Model-based control

Usually math-models tend to be an assumption (reducedorder model) and the crane motor behavior is far less linear than assumed in the model. The crane head only moves with friction. Disturbances such as wind cannot be modelled easily.

A Linguistic Control Strategy

A skilled operator is capable to control the crane. He does not even need to use differential equations or a cablelength sensor which many control techniques would require. So how does he do it?

Human-operated Crane System

Once he has picked the container, he starts the crane with medium power to see how the container sways. Depending on the reaction, he adjusts motor power to get the container a little behind the crane head.

In this position, maximum speed can be reached with minimum sway.

Getting closer to the target position, the operator reduces motor power or might even apply negative brake. With that the container gets a little ahead of the crane head until the container reaches the target position. Then motor power is increased so that the crane head is over the target position and sway is zero.

Analysis of Operators actions

1. Start with medium power.

2. If you get started and still far away from the target, adjust the motor power so the container gets a little behind the crane head.
3. If you are closer to the target, reduce motor speed so the container gets a little ahead of the crane head. 4. When the container is very close to the target position, power up the motor. 5. When the container is over the target and sway is zero, stop the motor.

See if you can write the rules to control this container crane system
First identify the antecedent variables Next the consequent variable Then write the rules according to the analysis of the operators action in the previous page. 6 rules can be written- Try?

The Control Strategy

1. 2. 3. 4. 5. 6. IF Distance = far AND Angle = zero THEN power = pos_medium IF Distance = far AND Angle = neg_small THEN power = pos_big IF Distance = far AND Angle = neg_big THEN power = pos_medium IF Distance = medium AND Angle = neg_small THEN power = neg_medium IF Distance = close AND Angle = pos_small THEN power = pos_medium IF Distance = zero AND Angle = zero THEN power = zero

Fuzzy Controller Design

From what you have studied thus far, lets design our Fuzzy Controller to solve this problem. What next?

Consequent

Conventional Fuzzy Control

Defuzzfication

Inference

Fuzzification

Antecedents

Antecedents

Partition or break your antecedents into several fuzzy sets that can reflect the system

 For each antecedent, identify the range for the universe of discourse. Distance Metres or Yards Angle From -90o to +90o Break up each antecedent 5 fuzzy sets each and provide the appropriate label that reflect the variables

Distance

Angle

too far

far
zero close medium

neg_big neg_ small zero pos_ big pos_ small

Next design appropriate membership functions for each fuzzy set and set them on the universe of each antecedent
Distance Angle

Typical design would be as follows:

Similarly for the consequent

Identify the motor power range Break up into 5 fuzzy sets
Power

neg_high pos_high neg_medium zero

pos_

medium

 Membership functions of the Consequent Motor Power

Next develop the rules Distance use matrix form

Too far
NB NS zero close med far

How many rules maximum?

Angle
ZE PS PB

Rules proposed by Fuzzy Tech software

Inference procedure?
Max-min or (min/max as described in FuzzyTech) Max-dot Etc.

Defuzzification
Centroid Mean of max

Try out the simulation exercise