- Title
- An evolutionary algorithm approach to ecological optimal control problems
- Creator
- Westcott, Amy-Rose
- Subject
- Mathematical optimization -- South africa
- Subject
- Evolutionary programming (Computer science)
- Date Issued
- 2022-04
- Date
- 2022-04
- Type
- Master's theses
- Type
- text
- Identifier
- http://hdl.handle.net/10948/55544
- Identifier
- vital:52812
- Description
- There are several challenges associated with applying conventional (hereafter classic) optimal control (OC) methods to ecological optimal control problems (OCPs). Conditions required by these methods, including differentiability and convexity, for example, are not always met, and ecological problems do not always adhere to solvable OCP formulations. Moreover, mathematically optimal solutions do not always translate to optimal ecological strategies in practice. Despite this, alternative OC approaches are relatively under-explored. Evolutionary algorithms (EAs) circumvent many of the complex aspects of classic OC methods and have been successfully applied to diverse OCPs. Nevertheless, EAs have sel dom been applied to ecological OCPs. The viability of an EA approach to ecological OCPs was therefore investigated in the current study, facilitated by four case studies of increasing complexity and a genetic algorithm (GA) as a representative EA approach. To ascertain the accuracy of a GA approach, comparisons between a GA and classic OC methods were conducted in the first three case studies. The GA generated near-optima in these comparisons, comparable to the corresponding classical solutions, whilst avoiding non-trivial mathematical theory. Supported by these results, an unconventional OCP, that arguably cannot be solved using classic OC methods, was formulated in the fourth case study, and solved using a GA approach. The resulting solution was feasible and further conformed with strategies found to be successful in practice. Additionally, the GA approach was rela tively simple to apply in all case studies. These collective outcomes of demonstrated the viability of a GA as an OC method in eco logical OCPs, thereby supporting the use of an EA approach as an alternative to classic OC methods in ecological OCPs. The feasibility of an EA approach to atypical OCPs was further demonstrated, which may act to increase realism in OC applications. Further investigation in this regard is thus warranted by this study.
- Description
- Thesis (MS) -- Faculty of Science, Mathematics and Applied Mathematics, 2022
- Format
- computer
- Format
- online resource
- Format
- application/pdf
- Format
- 1 online resource (xii,155 pages)
- Format
- Publisher
- Nelson Mandela University
- Publisher
- Faculty of Science
- Language
- English
- Rights
- Nelson Mandela University
- Rights
- All Rights Reserved
- Rights
- Open Access
- Hits: 606
- Visitors: 591
- Downloads: 48
Collections
NMU School of Computer Science, Mathematics, Physics and Statistics
| Thumbnail | File | Description | Size | Format | |||
|---|---|---|---|---|---|---|---|
| View Details Download | SOURCE1 | Westcott, AR.pdf | 4 MB | Adobe Acrobat PDF | View Details Download |