Show simple item record

dc.contributor.advisorSaha, Akshay Kumar.
dc.contributor.advisorModi, Albert Thembinkosi.
dc.creatorOyebode, Oluwaseun Kunle.
dc.date.accessioned2020-09-17T14:43:53Z
dc.date.available2020-09-17T14:43:53Z
dc.date.created2020
dc.date.issued2020
dc.identifier.urihttps://researchspace.ukzn.ac.za/handle/10413/18677
dc.descriptionDoctoral Degree. University of KwaZulu-Natal. Durban.en_US
dc.description.abstractThis study presents the development of a differential evolution (DE)-inspired artificial neural network (ANN) that incorporates climate and socioeconomic information for a more accurate and reliable water demand forecasting. The study also addresses the limitations of ANN. Multiple feature selection techniques were employed to identify the minimal subset of features for optimal learning. The performance of the feature selection techniques was validated and compared to a baseline scenario comprising a full set of data covering potential casual variables including weather, socio-economic and historical water consumption data. The performance of the models was evaluated based on accuracy. Results show that all the feature selection techniques outperformed the baseline scenario. More importantly, the subset of features obtained from the Pearson correlation technique produced the most superior model in terms of model accuracy. Findings from the study suggests that inclusion of weather and socioeconomic variables in water demand modelling could enhance the accuracy of forecasts and cater for the impacts of climate and socioeconomic variations in water demand planning and management. The performance of the optimal DE-inspired model was thereafter compared to those developed via conventionally-used multiple linear regression and standard time series technique – exponential smoothing as well as other prominent soft computing techniques, namely support vector machines (SVM) and conjugate-gradient (CG)-trained multilayer perceptron (MLP). Results show that the DE-inspired ANN model was superior to the four other techniques for both the baseline scenario and optimal subset of features. DE showcased robustness in fine-tuning algorithm parameter values thereby producing good performance in terms of the solution efficiency and quality. Generally, this study demonstrates that water demand models can account for the impacts of weather and socioeconomic variations by incorporating explanatory variables based on weather and socioeconomic factors. This study also suggests that the synergetic use of feature selection techniques, DE algorithm and an early stopping criterion could be used in addressing the limitations of ANN and developing an improved and more reliable water demand forecasting model. This work goes further to propose for a novel and more comprehensive integrated water demand and management modelling framework (IWDMMF) that is capable of syncing conventional evolutionary computation techniques and social aspects of society. The methodologies, principles and techniques behind this study fosters sustainable development and thus could be adopted in planning and management of water resources.en_US
dc.language.isoenen_US
dc.subject.otherArtificial neural network.en_US
dc.subject.otherMunicipal water demand.en_US
dc.subject.otherSustainable development.en_US
dc.subject.otherWater resources management.en_US
dc.subject.otherWater demand forecasting.en_US
dc.subject.otherWater consumption.en_US
dc.titleDevelopment of a sustainable evolutionary-inspired artificial intelligent system for municipal water demand modelling.en_US
dc.typeThesisen_US
dc.description.notesPublications from this thesis can be found on page v.en_US


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record