Intelligent sizing and output prediction techniques for grid-connected photovoltaic system / Shahril Irwan Sulaiman

This thesis presents new intelligent-based techniques for sizing and output prediction in grid-connected photovoltaic (GCPV) system. Initially, two intuitive-based sizing algorithms for GCPV system design, termed as Conventional Sizing Algorithm (CSA), were developed based on two design goals. The design goal 1 (DG1) was formulated with the aim to design a system based on a specific solar electricity requirement while the design goal 2 (DG2) dealt with the sizing of the system such that maximum solar electricity can be generated using the available roof space. In addition, each CSA incorporated both technical and economic sizing procedures to provide comprehensive design using a pre-selected set of photovoltaic (PV) module and inverter. It was found that the maximum solar electricity requirement from CSA with DG1 was actually capped to the maximum solar electricity that can be generated from CSA with DG2. Apart from that, a novel Iterative Sizing Algorithm (ISA) was also developed to present an iterative approach towards sizing process when there were numerous sets of PV module and inverter need to be considered. At this stage, a database of PV module and a database of inverter were developed to form possible sizing solutions.