UUM Electronic Theses and Dissertation
UUM ETD | Universiti Utara Malaysian Electronic Theses and Dissertation
FAQs | Feedback | Search Tips | Sitemap

Enhanced evolutionary algorithm with cuckoo search for nurse scheduling and rescheduling problem

Lim, Huai Tein (2015) Enhanced evolutionary algorithm with cuckoo search for nurse scheduling and rescheduling problem. PhD. thesis, Universiti Utara Malaysia.

[thumbnail of depositpermission_s91515.pdf] Text
depositpermission_s91515.pdf
Restricted to Repository staff only

Download (240kB)
[thumbnail of s91515_01.pdf]
Preview
Text
s91515_01.pdf

Download (12MB) | Preview

Abstract

Nurse shortage, uncertain absenteeism and stress are the constituents of an unhealthy working environment in a hospital. These matters have impact on nurses' social lives and medication errors that threaten patients' safety, which lead to nurse turnover and low quality service. To address some of the issues, utilizing the existing nurses through an effective work schedule is the best alternative. However, there exists a problem of creating undesirable and non-stable nurse schedules for nurses' shift work. Thus, this research attempts to overcome these challenges by integrating components of a nurse scheduling and rescheduling problem which have normally been addressed separately in previous studies. However, when impromptu schedule changes are required and certain numbers of constraints need to be satisfied, there is a lack of flexibility element in most of scheduling and rescheduling approaches. By embedding the element, this gives a potential platform for enhancing the Evolutionary Algorithm (EA) which has been identified as the solution approach. Therefore, to minimize the constraint violations and make little but attentive changes
to a postulated schedule during a disruption, an integrated model of EA with Cuckoo Search (CS) is proposed. A concept of restriction enzyme is adapted in the CS. A total of 11 EA model variants were constructed with three new parent selections, two new crossovers, and a crossover-based retrieval operator, that specifically are theoretical contributions. The proposed EA with Discovery Rate Tournament and Cuckoo Search Restriction Enzyme Point Crossover (DᵣT_CSREP) model emerges
as the most effective in producing 100% feasible schedules with the minimum penalty value. Moreover, all tested disruptions were solved successfully through preretrieval and Cuckoo Search Restriction Enzyme Point Retrieval (CSREPᵣ) operators.
Consequently, the EA model is able to fulfill nurses' preferences, offer fair on-call delegation, better quality of shift changes for retrieval, and comprehension on the two-way dependency between scheduling and rescheduling by examining the
seriousness of disruptions.

Item Type: Thesis (PhD.)
Supervisor : Ramli, Razamin
Item ID: 5794
Uncontrolled Keywords: Hybrid evolutionary algorithm, Cuckoo search, Restriction enzyme, Nurse scheduling and rescheduling problem (NSRP), Healthcare management.
Subjects: Q Science > QA Mathematics
Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Divisions: Awang Had Salleh Graduate School of Arts & Sciences
Date Deposited: 01 Aug 2016 17:08
Last Modified: 18 Mar 2021 08:31
Department: Awang Had Salleh Graduate School of Arts and Sciences
Name: Ramli, Razamin
URI: https://etd.uum.edu.my/id/eprint/5794

Actions (login required)

View Item
View Item