Date of Award


Degree Type


Degree Name

Doctor of Philosophy


School of Engineering


Faculty of Health, Engineering and Science

First Advisor

Dr Ganesh Kothapalli

Second Advisor

Dr Yasir Al-Abdeli


Wind and solar irradiance are promising renewable alternatives to fossil fuels due to their availability and topological advantages for local power generation. However, their intermittent and unpredictable nature limits their integration into energy markets. Fortunately, these disadvantages can be partially overcome by using them in combination with energy storage and back-up units. However, the increased complexity of such systems relative to single energy systems makes an optimal sizing method and appropriate Power Management Strategy (PMS) research priorities.

This thesis contributes to the design and integration of stand-alone hybrid renewable energy systems by proposing methodologies to optimise the sizing and operation of hydrogen-based systems. These include using intelligent techniques such as Genetic Algorithm (GA), Particle Swarm Optimisation (PSO) and Neural Networks (NNs). Three design aspects: component sizing, renewables forecasting, and operation coordination, have been investigated. The thesis includes a series of four journal articles.

The first article introduced a multi-objective sizing methodology to optimise standalone, hydrogen-based systems using GA. The sizing method was developed to calculate the optimum capacities of system components that underpin appropriate compromise between investment, renewables penetration and environmental footprint. The system reliability was assessed using the Loss of Power Supply Probability (LPSP) for which a novel modification was introduced to account for load losses during transient start-up times for the back-ups.

The second article investigated the factors that may influence the accuracy of NNs when applied to forecasting short-term renewable energy. That study involved two NNs: Feedforward, and Radial Basis Function in an investigation of the effect of the type, span and resolution of training data, and the length of training pattern, on shortterm wind speed prediction accuracy. The impact of forecasting error on estimating the available wind power was also evaluated for a commercially available wind turbine.

The third article experimentally validated the concept of a NN-based (predictive) PMS. A lab-scale (stand-alone) hybrid energy system, which consisted of: an emulated renewable power source, battery bank, and hydrogen fuel cell coupled with metal hydride storage, satisfied the dynamic load demand. The overall power flow of the constructed system was controlled by a NN-based PMS which was implemented using MATLAB and LabVIEW software. The effects of several control parameters, which are either hardware dependent or affect the predictive algorithm, on system performance was investigated under the predictive PMS, this was benchmarked against a rulebased (non-intelligent) strategy.

The fourth article investigated the potential impact of NN-based PMS on the economic and operational characteristics of such hybrid systems. That study benchmarked a rule-based PMS to its (predictive) counterpart. In addition, the effect of real-time fuel cell optimisation using PSO, when applied in the context of predictive PMS was also investigated. The comparative analysis was based on deriving the cost of energy, life cycle emissions, renewables penetration, and duty cycles of fuel cell and electrolyser units. The effects of other parameters such the LPSP level, prediction accuracy were also investigated.

The developed techniques outperformed traditional approaches by drawing upon complex artificial intelligence models. The research could underpin cost-effective, reliable power supplies to remote communities as well as reducing the dependence on fossil fuels and the associated environmental footprint.