Integer programming approach for scheduling multi-stage multi-product flow-shop /
To face the challenges of industrial globalization and sustain in the competitive market, the manufacturers have to gratify the customer demand by launching the products on time having various design and volume at low price. In this regard, the necessity of adopting the epitome of flexible mass prod...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
Kuala Lumpur :
Kulliyyah of Engineering, International Islamic University Malaysia,
2019
|
| Subjects: | |
| Online Access: | http://studentrepo.iium.edu.my/handle/123456789/4896 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | To face the challenges of industrial globalization and sustain in the competitive market, the manufacturers have to gratify the customer demand by launching the products on time having various design and volume at low price. In this regard, the necessity of adopting the epitome of flexible mass production flow shop structure along with the appropriate production planning tools and techniques like scheduling knows no bound. As a consequence, numerous approaches have already proposed for scheduling the production flow shop in order to find the most optimize solution. This problem is known as NP-hard and no approaches that can guarantee the optimality of the solution. However, before the adoption of any of these conventional approaches it is an utmost need for the manufacturer to realize its consequences and the appropriateness. Previous approaches also not highlighting the solution considering the combination of multi-stage flow shop with various number of parallel machines and multi-product environment. They only emphasized the models of hybrid flow-shop in two or three stages which is in a limited scope compared to a real industrial scenario. Therefore, this research study is aiming to develop a mathematical model of machine scheduling with unrelated parallel machines and shared resources in flow-shop environment and minimizing the makespan by finding the best sequence of scheduling. In this endeavour, this study anticipated Mixed Integer Non-Linear Programming (MINLP) models for machine scheduling in flow shop environment based on multi-stage and multi product production based on Wagner's. A case study has been conducted in a local concrete pole production company and from the data obtained, two models were developed. The first model is on general flow-shop of 13 stages with single machine per stage while the second model is on 13 stages flow-shop with maximum number of three unrelated parallel machines per stage. Each of the model has been analysed using What's Best Excel Solver. The result shown by adopting the proposed sequence, it is possible to achieve the minimum completion time. The proposed sequence for general flow-shop model is 1,2,3 with estimated production makespan of 1822.13 seconds and the proposed sequence for multi-stage multi-product flow-shop is 3,1,2 with overall makespan of 2019.96 seconds in order to completely produced three products in one cycle. The second model obtained lower performance of sequence decision but produced moderate result in term of resource utilization which has 0.29 percent differences with the job sequence that has the highest mean utilization rate. By minimizing the makespan, the idle times of some of the machine were reduced meanwhile the utilization of the machine were maximized consequently. Due to large number of decision variables that need to be considered to represent the flow-shop complexities, hence, in future research, there is a need to improvise the solution so that the approach can adapt with enormous volume of decision variables and constraints. Apparently, parameters such as setup time, loading and unloading time should be treated and some suggestion for modification can be made in the production line due to waiting time resulted from the proposed approach. |
|---|---|
| Physical Description: | xvi, 121 leaves : colour illustrations ; 30cm. |
| Bibliography: | Includes bibliographical references (leaves 110-117). |