Optimizing driver scheduling for UTeM shuttle bus using harmony search
Scheduling is one of the decision-making forms that play a vital role in manufacturing and service industries. The main problem in scheduling is fairness, so a good schedule is the key to maintain the effectiveness of an operation. Scheduling is very important when dealing with task distributions an...
Saved in:
| 总结: | Scheduling is one of the decision-making forms that play a vital role in manufacturing and service industries. The main problem in scheduling is fairness, so a good schedule is the key to maintain the effectiveness of an operation. Scheduling is very important when dealing with task distributions and time management. The important tasks can be covered at appropriate times with the existence of proper scheduling. It will be more productive; well organized and manageable. There are many fields that required the scheduling such as transportation (bus, train and flight scheduling), medical field (nurse scheduling), manufacturing (production and workers’ shift scheduling) and education (course and examination scheduling). This project is focus more on driver scheduling for UTeM shuttle bus. Driver scheduling (DSP) can be defined as the process of assigning shift and route to driver according to the bus schedule that
has been provided over a scheduling period. Bahagian Pengurusan Kenderaan Universiti (BPKU) is one of the departments in UTeM which plays the role to organize the driver schedule every month manually. The process of generating the schedule is complicated since the shift and route given to drivers should be balance based on some constraints. Therefore, this project is carried out to generate an optimized schedule automatically using Harmony Search (HS). HS is one of the new optimization techniques that already solved many optimization problems. For DSP, all data are collected during interview session with BPKU.
The problems arise in DSP in UTeM were analyzed and related to HS techniques. It is important to consider all hard constraints and soft constraints in order to produce a balance schedule. The DSP is implemented based on 5 steps of HS. The goal of this project is to minimize the objective function, which is minimizing the soft constraint violation. The result produced for this project is quite promising since the objective function obtained is better than real schedule which is done manually. The t-test is performed to compare these two samples. The value obtained is less than 0.05, so, there is a significant difference between the means of
these two samples. |
|---|