A large number of power sources, operational in a microgrid, optimum power sharing and accordingly controlling the power sources along with scheduling loads are the biggest challenges in modern power system. In the era of smart grid, the solution is certainly not simple paralleling. Hence it is required to develop a control scheme that delivers the overall power requirements while also adhering to the power limitations of each source. As the penetration of distributed generators increase and are diversifi ed, the choice of decentralized control becomes preferable. In this work, a decentralized control framework is conceived. The primary approach is taken where a small hybrid system is investigated and decentralized control schemes were developed and subsequently tested in a hardware in the loop in conjunction with the hybrid power system setup developed at the laboratory. The control design approach is based on the energy conservation principle. However, considering the vastness of the real power network and its complexity of operation along with the growing demand of smarter grid operations, called for a revamp in the control framework design. Hence, in the later phase of this work, a novel framework is developed based on the coupled dynamical system theory, where each control node corresponds to one distributed generator connected to the microgrid. The coupling topology and coupling strengths of individual nodes are designed to be adjustable. The layer is modeled as a set of coupled differential equations of pre-assigned order. The control scheme adjusts the coupling weights so that steady state constraints are met at the system level, while allowing flexibility to explore the solution space. Additionally, the approach guarantees stable equilibria during power redistribution. The theoretical development is verified using simulations in matlab simulink environment.

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.

No abstract provided. / Thesis / Master of Engineering (MEngr)

Silicon has been the mainstay of the microelectroncs industry for over four decades. There is no material which can match the balance it affords between cost-benefit, mass consumability, process versatility, and nano-scale electron device performance. It is, therefore, the logical (and perhaps inevitable) platform for the development of integrated opto-electronics - a technology that is being aggressively developed to meet the next generation of bandwidth demands that are already beginning to strain interconnect architectures all the way down to the intra-chip level. While silicon-based materials already provide a variety of passive optical functionalities, the success of a genuine silicon-based optoelectronics will depend upon the ability of engineers to overcome those limitations in the optical properties of bulk silicon that occur at critical junctions in device requirements (eg. modulator and laser). Such solutions must not render the device processing incompatible with CMOS, for then the "silicon advantage" is lost. Achieving reliable and efficient electroluminescence in silicon remains the most intractable of these problems to date. Reliability problems in recently developed light emitting devices operating near a wavelength of 1.54 f..Lm, based on the thermally induced formation of silicon nano-clusters in erbium-doped silicon rich silicon oxide thin films, has reinforced the need for a further understanding of the luminescence mechanisms in this material. Indeed, the efficient and stable sensitized photoluminescence from Er3+ ions (near the telecom wavelength), embedded in an oxide matrix, based on a quasi-resonant energy transfer from nanostructured silicon, has the potential to make possible compact waveguide amplifiers and thin film electroluminescence. This thesis represents a study into the luminescence mechanisms in erbium-doped silicon oxide (SiOx, x~2) thin films grown by electron cyclotron resonance plasma enhanced chemical vapour deposition. Importantly, the film growth relies on in-situ erbium doping through the cracking of a volatile organalanthanide Er(tmhd)3 source. Rutherford backscattering spectroscopy has been used to map the film composition space generated from an ECR-PECVD parameter subspace consisting of precursor gas flow rates and the erbium precursor temperature. The response of the film photoluminescence spectra in both visible and infrared bands consistenly reveals three classes of luminescence centres, whose relative ability to emit light is shown in this study to exhibit a considerable degree of variability through the control of the film composition, subsequent thermal anneal temperature, duration, and process ambient. These three classes consist of optically active Er3 + ions, silicon nano-clusters phase separated during thermal annealing, and oxide-based defects (which may additionally include organic chromophores). The latter two of these species show the ability to sensitize the Er3 + luminescence. In fact, sensitization by intrinsically luminescent defects is a rarely studied phenomenon, which seems to be an important phenomenon in the present films owing to a potentially unique Er incorporation complex. To further investigate the ability of the oxide defects in this regard, an optimally luminescent film has been subject to a damaging ion irradiation to induce a photoluminescence quenching. The subsequent recovery of this luminescence with stepwise isochronous annealing has been correlated with Doppler broadening positron annihilation spectroscopy measurements made with a slow positron beam. Irradiation to a sufficiently high fluence has demonstrated a unique ability to de-couple luminescent sensitizers and Er3+ ions, producing enhanced blue and violet emissions. / Thesis / Master of Science (MS)

Metal-semiconductor-metal (MSM) photodetectors are becoming attractive devices for optoelectronic integrated circuits due to their high speed and simplicity. Optoelectronic matrix switches based on MSM detector arrays offer many advantages such as zero-bias off-state, low bias voltage, high speed and large bandwidth. While in many applications the optical input is coupled in through the top surface of the device, optical signals can also be distributed through transparent waveguides that are located below the absorbing detector layer. Such waveguide-coupled detectors will act as optical taps when the coupling between the waveguide and detector layers is well under control. In this thesis, a 4x4 MSM waveguide-coupled optoelectronic matrix switch was demonstrated and analyzed. The strength of the coupling between the waveguide and detector layers was predicted theoretically and confirmed experimentally. Franz-Keldysh effect in this device was also demonstrated. / Thesis / Master of Engineering (ME)