The application of fuzzy expert system to preliminary development planning of medium size container terminal
The aim of this study is to develop a fuzzy expert system to serve as an alternative to the conventional method of container terminal development planning. The conventional method commonly used by port planners in developing countries employs planning charts and empirical formulas. It requires preci...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2006
|
| Subjects: | |
| Online Access: | http://eprints.utm.my/id/eprint/1299/1/MohdZamaniAmhadPBATC2006.pdf http://eprints.utm.my/id/eprint/1299/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The aim of this study is to develop a fuzzy expert system to serve as an alternative to the conventional method of container terminal development planning. The conventional method commonly used by port planners in developing countries employs planning charts and empirical formulas. It requires precise inputs although these values are forecasted and imprecise. It is short of component which allows planning under uncertainty. Hence, there is a real need to improve the current method so that port planners in developing countries can apply their natural modes of reasoning that involves approximate, imprecise, linguistic and subjective values. The study has proposed an enhanced model of container terminal development planning. It has applied triangular membership functions concept and centre of gravity method for the fuzzification and defuzzification processes respectively. Fuzzy Associative Memory (FAM) method has been used to derive rules for the CLIPS (C Language Integrated Production System) expert system database. The system developed has been verified against the conventional method to confirm its accuracy and a case study has been performed to prove its practical usability. Data extraction and expert knowledge generation involving fuzzification and defuzzification processes has been done accurately. The Pearson r-squared analysis performed on the correlation lines did not show any inconsistency in quality. More than three quarter of the rules used represent genuine expert knowledge while the rest are rules that store default values. The verification results show that the system is accurate and no indication of inefficiency from the use of forward chaining CLIPS. The case study proves that the expert system database is complete and all unexpected results are traceable to the inappropriate combination of planner inputs. Therefore, the study has successfully developed an accurate and efficient fuzzy expert system which can serve as an alternative tool for container terminal development planning and solves the problem of lack of human modes of reasoning found in the conventional methods |
|---|